## EXECUTIVE SUMMARY ## Source details **Canonical URL:** [EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/cr/2018/cr18228.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2018/cr18228.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2018/cr18228.pdf.json) --- ### Key solvency findings - Aggregate common equity tier 1 (CET1) ratio: - Starting point: 13.9 percent at the end of 2017. - Under the adverse scenario: low point of 10.0 percent in 2020. - Comparable results obtained using alternative solvency metrics. - Aggregate capital buffers are sizeable relative to immediate threats, but vulnerabilities are heterogeneous: - Some banks especially vulnerable to credit risk. - Others more exposed to market risks from risk premia decompression, basis risk, and valuation shocks. - Business risk from lower-than-expected volumes or interest margins relevant for some banks. - Relative resilience by bank type: - Large, but less complex, internationally-active banks: more robust to deterioration in economic and financial conditions. - G-SIBs: business models emphasizing trading and capital markets-related activities disproportionately affected by market and funding liquidity risks, especially in foreign currencies. - Domestically-oriented banks: more vulnerable to deteriorations in domestic macroeconomic conditions and, in some instances, to market-wide liquidity shocks. ### Liquidity resilience and vulnerabilities - System liquidity resilience has increased, but multiple banks may face liquidity challenges in extreme scenarios when assuming only marketable securities are available. - Eurosystem extraordinary operations changed: - Composition of banks’ liabilities. - Structure of their counterbalancing capacity (CBC) by increasing banks’ reliance on ECB support. - Banks differ notably in funding structures, dependence on foreign currency funding, asset encumbrance, and amount/composition of CBC. - G-SIBs: - More vulnerable to market runs due to funding structure and trading activities. - Generally have sufficient CBC to cope with funding liquidity shocks, but some may face challenges replacing central bank funding with market alternatives. - Banks most at risk: those more dependent on central bank refinancing operations and those with lower quality collateral facing higher funding costs when reverting to market funding. ### Solvency–liquidity interactions - Solvency and liquidity risks may feed on each other: - Abrupt and sharp increases in rates can materially impact bank capital as funding costs outpace repricing of assets and valuation effects erode buffers, particularly when fair value hedges are ineffective. - Funding outflows triggered by solvency concerns likely to result in some banks breaching capital requirements and facing negative CBC positions. - Solvency–funding loops are exacerbated by non-linear interactions among credit, liquidity, and market shocks. ### Main policy and supervisory implications - Enhance the ECB’s stress testing analytical framework by: - Further integrating supervisory data, especially related to market risk and cash flow based liquidity risks, into stress testing infrastructure — Timing: ST. - Deepening analysis of risk interactions to examine their role in aggravating systemic risk — Timing: MT. - Expanding the macroprudential stress testing framework to include significant subsidiaries where failure could cause major disruption and sufficient granular supervisory data is available — Timing: MT. - Data collection priorities: - Collect cash-flow data for all significant currencies — Timing: I. - More data needed to monitor market, funding, and contingent liquidity risks; collect cash flow data by significant jurisdictions to monitor availability and transferability of collateral. - Step up collection on securities financing transactions and derivatives to enhance monitoring of market and contingent liquidity risks, including flows of collateral via CCPs and cyber risks. - Main recommendations on stress testing (timing legend preserved verbatim): - Solvency Stress Testing: - Make further use of supervisory data for stress testing, enhancing analysis, and feeding back into supervision — Timing: ST. - Deepen analysis of banking group risks by extending the perimeter of macroprudential stress tests to cover systemic subsidiaries of EA significant institutions and significant investment firms — Timing: MT. - Continue enhancing the assessment of banks’ risk interactions across various risk categories — Timing: MT. - Liquidity Stress Testing: - Use cash-flow data for all significant currencies and significant jurisdictions to refine liquidity monitoring and stress testing — Timing: I. - Enhance collection of data on, and analysis and supervision of contingent liquidity risks associated with cross-border flow of collateral, securities funding transactions, and cyber risks — Timing: ST. - Timing legend: - I (immediate) = within one year - ST (short term) = 1–2 years - MT (medium term) = 3–5 years *Source: EXECUTIVE SUMMARY, cr18228.* --- ### 2. Euro area banks: size, trends, and stress-testing approach ### Overview and key statistics - Euro area banking sector total assets (consolidated, including foreign subsidiaries and branches): around €35 trillion in 2017. - Since the global financial crisis: shift toward safer and less complex assets and more stable funding sources; nonbanks and financial markets play a greater role in funding larger nonfinancial firms. ### FSAP stress-testing sample and coverage - Solvency stress test sample: 28 banks covering broadly 65 percent of banking system assets; banks account for over 15 trillion euros. - Liquidity stress test sample: 29 banks covering broadly 70 percent of banking system assets. - FSAP sample covers around 85 percent of the 2018 EBA sample. - Sample composition: - Six (seven) G-SIBs of the 30 G-SIBs identified by the FSB in November 2017 for the solvency (liquidity) test. - Banks grouped into three categories: G-SIBs; large but less complex internationally-active banks; relatively smaller domestically-focused banks. - Table-based aggregate asset subtotals (actual data end-2017) from the sample groups: Sub-total 8,910 (billions of euros); Sub-total 5,207 (billions of euros); Sub-total 2,433 (billions of euros). - G-SIB capital surcharge buckets in the sample: one G-SIB in bucket 3 (2.0 percent), one in bucket 2 (1.5 percent), and 4 in bucket 1 (1.0 percent). - Material geographies for the sample: 9 home euro area countries, 11 other EU countries, and 17 countries outside the EU (37 material geographies). ### FSAP stress-testing strategy and scenarios - Two-pronged approach: - Solvency stress tests: scenario-based test, sensitivity tests, and targeted reverse stress test (including valuation shocks on opaque complex assets). - Liquidity stress tests: structural liquidity analysis (LCR, NSFR, Asset Encumbrance, funding concentration) and cash-flow tests under 5-day, 1-month, and 3-month horizons. - Scenario characteristics: - FSAP adverse scenario features long-lived shocks (U-shape) over a three-year horizon Q1 2018–Q4 2020. - Implied real GDP growth rate in the euro area under FSAP adverse: -1.6 percent at year t+2 by 2019; FSAP posts -0.3 percent growth in 2020 under the scenario. - Overall adverse scenario: output contracts by 7.1 percent relative to the baseline by 2020; consumer price inflation falls by 3.3 percentage points; total employment falls by 5.9 percent in the euro area. - Market shocks include: equity prices falling by 20 percent in euro area/UK/US over two years; money market spreads widen by 75 bps in the euro area; term premium decompression raises long-term government bond yields by 150 bps in high-spread euro area economies and 75 bps in low-spread economies. --- ### Credit risk, market risk, and IRB/STA methodology ### Credit risk modeling approach and coverage - Aggregate figures: - Size of largest 28 banks: €15.0 trillion at end-2017; RWAs of €5.6 trillion (around 37 percent risk density). - Over 85 percent of capital required to cover unexpected losses from credit risk; market risk under 10 percent of capital requirements. - Parameter pool and projections: - Stressed projections applied to around 10,000 initial credit risk parameters (PD, LGD, EaD, RWA) broken down by 7 exposure classes, 12 geographies, and 28 banks. - Two-step PD projection: - Step 1: econometric methods linking economic/financial conditions to proxies on credit risk parameters. - Step 2: statistical adjustment to bank-specific portfolios drawing on post-June 2014 supervisory data. - Sovereign PD proxy: - PD proxy for central government exposures backed out from sovereign yields using a Merton-based reduced-form structural model assuming LGD=45 percent; implied risk-neutral PD: PD = 1 - exp(-Si,tT * (T-t) / LGD). - LGD and EAD: - LGD projections informed by banks’ reported projections and ECB staff multipliers; median LGD increases under adverse scenario: residential LGD up 50 percent; commercial LGD up 12 percent. - EAD drivers include credit growth, foreign currency exposure, triggered credit lines and guarantees; stressed credit conversion factors informed by historical off-balance sheet migration and Pillar 3 disclosures. - RWAs: - RWAs computed after applying a scaling factor of 1.06 to credit RWAs and using a 1.25 multiplier to the correlation parameter for exposures to large regulated and unregulated financial institutions. - PiT shifts applied to regulatory PDs for IRB formula. ### Market risk and valuation approach - Scope: - All positions under fair value measurement revalued: HFT, FV through P&L, trading financial assets, AFS (OCI impact). - Excludes amortized cost positions in a hedge-accounting relationship and hedge accounting derivatives; changes in CVA and CCR excluded. - Valuation/hedging assumptions: - Hedging instruments for interest rate risk in trading book and AFS portfolios assumed ineffective under stress. - Traded risk losses partially reversed as asset prices recover after a sharp correction assumed during the first year. - Shock application: - Instantaneous shock applied to derivatives and equity instruments in trading book; multi-year shocks applied to debt instrument valuations over the 3-year horizon. - Sovereign and corporate valuation specifics: - Average euro area bank holds around 75 percent of sovereign portfolio in AFS; around 90 percent of sovereign securities held at fair value (AFS and HFT). - Banks hold about 9 percent of total assets in sovereign debt securities (June 2017), of which 30 percent are to their own sovereign. - Corporate exposures at fair value can reach up to 15 percent of balance sheet for some banks; Moody’s Baa corporate bond yield rise by 180 bps under stress is an example calibration. ### Interest Rate Risk in the Banking Book (IRRBB) - Net interest margin (NIM) benchmarks: - Euro area NIM around 120 bps, compared with 200 bps (Australia), 230 bps (UK), 315 bps (US), 85 bps (Japan). - Swap curve flattening: - Swap curve slope fell from 180 bps in 2009 to 100 bps at end-2017. - Baseline expectation: - 6-month Euribor expected to increase by 100 bps by 2021 from -32 bps at end-2017 under the baseline. - FSAP IRRBB assessment: - Four-pronged: economic value (EVE) revaluation of marked-to-market assets; earnings (NII) over 3 years; scenario composition with systematic and idiosyncratic shocks; application to repricing structure using STE template. - Sensitivity: hypothetical 200 bps parallel upward shift reduces aggregate CET1 by 200 bps (single factor test). --- ### Stress test results: solvency, profitability, and single-factor sensitivities ### Solvency results (aggregate and drivers) - Adverse scenario aggregate impacts: - Aggregate risk-weighted CET1 capital ratios fall by about 390 basis points in the adverse scenario (with slightly larger impact on G-SIBs). - Loan impairment charges reduce aggregate CET1 ratio by about 3.0 percentage points. - Rising RWAs reduce aggregate CET1 ratio by around 2.1 percentage points. - Valuation losses from market price movements decrease CET1 ratio by 1.0 percentage point on average. - Interest rate risk on net income reduces CET1 by around ½ percentage point on average. - Pre-provision net revenue increases aggregate CET1 ratio by 3.1 percentage points relative to the starting point. - Profitability: - On aggregate, banks post negative net profits of €24 billion by 2020 from a pre-stress level of €51 billion in 2017. - Dispersion: - Coefficient of variation of CET1 increases from 16 percent in 2017 to 23 percent in 2020 under stress. ### Single-factor sensitivity and reverse tests - Interest rate shock: - 200 basis points parallel upward shift reduces aggregate CET1 by 200 bps. - Sovereign spreads: - 200 bps widening on own sovereign spreads lowers post-stress aggregate CET1 by 30 bps. - Corporate spread widening (2008-like): - Further 175 bps reduction in CET1; G-SIBs more severely affected. - Credit rating downgrade: - One-notch downgrade on standardized exposures erodes capital ratios by 60 bps; domestically-oriented banks impacted more. - LGD floor on mortgages: - LGD floor of 30 percent on residential mortgages reduces aggregate CET1 by around 40 bps. - Reverse sensitivity on hard-to-value assets: - 10 percent valuation shock in Level 3 assets and 5 percent in Level 2 assets could deplete capital buffers of some large banks below SREP regulatory minimum. ### Stress test insight on predictive power of indicators - Balance-sheet metrics and market-based indicators insufficient to predict banks’ stressed financial strength. - Lack of co-movement between initial metrics (CET1, NIM, NPLs, CDS spreads) and capital depletion under stress; higher initial capital buffers generally helpful but other metrics often do not co-move with stressed outcomes. --- ### Liquidity tests: cash-flow analysis, CBC, and scenarios ### Cash-flow Liquidity Stress Tests (CFLST) — data and indicators - Perimeter: 29 banks, consolidated; baseline date September 30, 2017. - Key indicators: - Net-funding gap (NFG): cash-inflows minus cash-outflows per bucket; cumulated net-funding gap (CNFG). - Counterbalancing capacity (CBC): cash inflows banks can generate under stress at reasonable prices in the bucket. - Reported contractual liquidity exposure: - Contractual outflows within first 4 weeks: about 31 percent of total assets (TA) excluding open maturity/overnight retail and corporate deposits (which are 23 percent and 7.5 percent of TA respectively). - Contractual inflows: about 21 percent of TA (excluding central bank deposits due to reporting). - Cumulated net funding gap over first 4 weeks: about 10 percent of TA or €1,700 billion. - Gross encumbrance due to repos: about 14 percent of TA; gross reverse repos: about 10 percent of TA. - CBC size and composition: - System-wide CBC in first month fully covers system-wide cash flow gap; CBC equals about 20 percent of TA. - CBC composition (weighted average as percent of TA, 2017 Q3): - Central bank deposits: 7.3 - 0 percent risk-weight securities: 5.8 - Other CB eligible assets (credit claims): 3.4 - Covered bonds: 0.8 - RMBS: 0.5 - Cash: 0.3 - Average remaining maturity of sovereign securities in CBC close to two years. ### Scenario set and CBC approaches - Scenario set: 20 embedded scenarios for 4-week and 3-month horizons and five scenarios for 5-day horizon: baseline plus mild market, medium market, severe market, combined severe market-wide and idiosyncratic stress. - Four CBC approaches: - (i) full CBC without haircuts, - (ii) full CBC with bank-specific haircuts, - (iii) marketable CBC (disregard non-marketable CBC components), - (iv) liquid CBC (bank-specific haircuts and market price effects from solvency adverse scenario). - Haircuts: ECB average haircuts reported (selected): - 0 percent RW securities: 4.2 (average haircut across banks). - Covered bonds: 7.1. - Other central bank eligible assets (credit claims): 26.1. - Equities: 25.0–75.0. - Undrawn committed lines provided to the bank: 50.0. ### CFLST results — 4-week horizon (selected outcomes) - Aggregate impacts in percent of TA (Table 9 averages): - Full CBC w/o HCs: Mild Market -5; Medium Market -8; Severe Market -12; Severe M/Idiosyncratic -14. - Full CBC w HCs: Mild Market -4; Medium Market -7; Severe Market -11; Severe M/Idiosyncratic -13. - Marketable CBC: Mild Market -7; Medium Market -10; Severe Market -14; Severe M/Idiosyncratic -16. - Liquid CBC: Mild Market -7; Medium Market -10; Severe Market -14; Severe M/Idiosyncratic -16. - Number of banks with negative CCBC (4-week, Liquid CBC, Severe M/Idiosyncratic): 6 banks. - Total contribution to decline in CBC if only marketable collateral used: €136 billion, or 8 percent of initial CBC. - In most severe case: total decline in CBC close to 20 percent of initial CBC. ### CFLST results — 3-month horizon (selected outcomes) - Aggregate impacts in percent of TA (Table 12 averages): - Full CBC w/o HCs: Severe M/Idiosyncratic -16. - Marketable CBC / Liquid CBC: Severe M/Idiosyncratic -18. - Under Liquid CBC severe/idiosyncratic 3-month scenario: - Aggregate CBC drops from €3.2 trillion to €45 billion. - Number of banks with negative CCBC: up to 11 banks under Marketable/Liquid CBC depending on scenario severity. - Aggregate shortfall up to 1.3 percent of total assets for most severe 3-month scenario when disregarding non-marketable CBC. ### Special 5-day collateral freeze (cyber-risk) scenario - Assumption: collateral held at CCPs and available for rehypothecation inaccessible for five business days. - Results: - In baseline: no bank’s liquidity buffers depleted. - In adverse/severe: a few banks fail the test; total impact on CBC from event: 9 percent decline. - G-SIBs and internationally-active banks more affected due to rehypothecation reliance. ### Key liquidity drivers and vulnerabilities - Major drivers of stress: net unsecured deposits of financial institutions and net repos. - Bank types most vulnerable: - G-SIBs: higher outflows due to short-term wholesale funding reliance. - Some domestic banks: lower CBC buffers, higher asset encumbrance, reliance on central bank funding. - Asset encumbrance and CBC heterogeneity: - Relative encumbrance ratio average in sample: 65 percent. - Average ability to encumber ratio across 29 banks: 35 percent. - Contingent liquidity: €738 billion of additional collateral would be needed if collateral values decline by 30 percent (23 percent increase in aggregate encumbrance). --- ### Funding costs, feedback loops, and integrated solvency–liquidity analysis ### Adverse scenario impact on funding costs - Deposit rates: rise by 15 percent, 42 percent, and 56 percent over 2018–2020 relative to starting point. - Debt spreads: increase by up to 240 percent by 2020 relative to starting point. - Funding costs increase unevenly; wholesale funding more sensitive than retail deposits. ### Exploratory solvency–liquidity feedback (Module 1) - Iterative quarterly process 2018Q1–2020Q4 linking funding costs, lending pass-through, and CET1. - Econometric findings: - A 100-bps increase in regulatory Tier 1 capital linked to a 20 bps decrease in effective customer deposit rate; relationship convex. - Pass-through coefficient of funding costs to lending rates estimated at 0.65 (statistically significant at 99 percent). - Iteration multipliers (median deposit-rate multipliers across banks): - Iteration 1: 2018 = 1.14, 2019 = 1.38, 2020 = 1.58. - Iteration 3 (converged): 2018 = 1.37, 2019 = 1.84, 2020 = 2.22. - Impact on capital: - Accounting for funding-cost feedback loops adds an average 30 bps to capital depletion; at the 90th percentile impact increases to 50 bps. ### Credit ratings, backbook effect, and Module 3 findings - Merton-based DRSK model used to map solvency stress inputs to implied default probabilities and ratings. - Backbook effect: - Stylized funding cost increases: Y1 +10 bps; Y2 +30 bps; Y3 +60 bps on repricing liabilities. - Average backbook effect equals about two-thirds of the entire funding cost effect. - Moderate average increase in funding costs implies an average CET1 impact of 88 bps after three years. - Econometric drivers of backbook effect (Table 15): - Risk density (RD = RWA/EAD) coefficient = -1.49551***. - Asset-liability mismatch (ALM) coefficient = 0.012794***. - Adjusted R-squared = 0.608477. - Heterogeneity: feedback effect higher for banks with higher ALM, higher unsecured wholesale funding share, lower risk density, lower repriceable loan volume, and less pricing power. --- ### Key quantitative magnitudes and metrics (selected) - Euro area banking sector assets (2017 consolidated): around €35 trillion. - Sample of 28 banks size: €15.0 trillion; RWAs €5.6 trillion (around 37 percent risk density). - CET1: start 13.9 percent (end-2017); adverse low 10.0 percent in 2020. - Aggregate CET1 reduction in adverse scenario: about 390 basis points. - Aggregate net profits: €51 billion in 2017 (pre-stress); projected net losses €24 billion by 2020 under adverse. - CBC and liquidity: - System-wide CBC before stress: €3.2 trillion. - Under 3-month severe/idiosyncratic liquid CBC scenario: aggregate CBC falls to €45 billion. - Cumulated net funding gap over first 4 weeks: about 10 percent of TA or €1,700 billion. - QE and TLTRO magnitudes: - QE: about €2,300 billion; outstanding TLTROs: around €750 billion (December 2017). - PSPP: €1,900 billion; CBPP: €240 billion; CSPP: €131 billion; ABSPP: €25 billion. - Haircuts (selected averages across banks): - 0 percent RW securities: 4.2 percent. - Covered bonds: 7.1 percent. - Other central bank eligible assets (credit claims): 26.1 percent. - Equity instruments: 25.0–75.0 percent. - Undrawn committed lines to the bank: 50.0 percent. --- ### Supervisory priorities and recommended bank actions ### Supervisory priorities - Prioritize credit risk in high-risk portfolios; step up prudential provisioning where needed. - Prioritize valuation effects from sudden asset-price corrections, basis widening, and hard-to-value asset shocks. - Focus supervisory attention on banks with negative post-stress CBC and those combining low capitalization with low liquidity. - Deepen monitoring of contingent liquidity risks from SFTs, derivatives, CCPs, and rehypothecation chains; collect more granular data. ### Bank-level actions - Strengthen balance sheets to ensure reasonable funding costs as reliance on private markets increases: - Lengthen funding tenors. - Increase liquidity buffers and CBC where needed. - Stagger tenors of deposits from financial institutions. - Improve risk-sensitivity of pricing of committed lines to customers. - Adjust business models and income diversification to mitigate compressed NII (e.g., increase fee & commission income), noting competition and stress-sensitivity of F&C. *Source: cr18228 — Chapter excerpts and executive summary.* ### EXECUTIVE SUMMARY ___________________________________________________________________________ 8 ### EXECUTIVE SUMMARY ### Key findings on solvency - The FSAP team undertook a thorough top-down stress testing analysis using end-2017 data covering scenario-based solvency tests, single factor sensitivity tests, and liquidity tests for a sample of major euro area banks supervised by the Single Supervisory Mechanism (SSM). - Aggregate common equity tier 1 (CET1) ratio: - Starting point: 13.9 percent at the end of 2017. - Under the adverse scenario: low point of 10.0 percent in 2020. - Comparable results are obtained using alternative solvency metrics. - Aggregate capital buffers are sizeable relative to immediate threats, but vulnerabilities are heterogeneous: - Some banks are especially vulnerable to credit risk. - Others are more exposed to market risks from risk premia decompression, basis risk, and valuation shocks. - Business risk from lower-than-expected volumes or interest margins is relevant for some banks. - Relative resilience by bank type: - Large, but less complex, internationally-active banks are more robust to deterioration in economic and financial conditions. - G-SIBs’ business models emphasizing trading and capital markets-related activities are disproportionately affected by market and funding liquidity risks, especially in foreign currencies. - Domestically-oriented banks are more vulnerable to deteriorations in domestic macroeconomic conditions and, in some instances, to market-wide liquidity shocks. ### Liquidity resilience and vulnerabilities - System liquidity resilience has increased in recent years, but multiple banks may face liquidity challenges in extreme scenarios when assuming only marketable securities are available. - EuroSystem extraordinary operations have changed: - Composition of banks’ liabilities. - Structure of their counterbalancing capacity (CBC) by increasing banks’ reliance on ECB support. - Banks differ notably in: - Funding structures. - Dependence on foreign currency funding. - Asset encumbrance. - Amount and composition of CBC. - G-SIBs: - More vulnerable to market runs due to funding structure and trading activities. - Generally have sufficient CBC to cope with funding liquidity shocks. - Some banks may face challenges in replacing central bank funding with market alternatives. - Certain banks that are more dependent on central bank refinancing operations and those with lower quality collateral are most at risk of facing higher funding costs. ### Solvency–liquidity interactions - The analysis suggests solvency and liquidity risk may feed on each other: - Sensitivity tests show an abrupt and sharp increase in rates can materially impact bank capital as funding costs outpace the repricing of assets and valuation effects erode buffers, particularly when fair value hedges are ineffective. - Funding outflows triggered by solvency concerns are likely to result in some banks breaching their capital requirements and facing negative CBC positions. - Solvency–funding loops are exacerbated by non-linear interactions among credit, liquidity, and market shocks. ### Policy and supervisory implications - The ECB’s stress testing analytical framework—already very sophisticated—could be enhanced along several dimensions: - Further integrating supervisory data, especially related to market risk and cash flow based liquidity risks, into the existing stress testing infrastructure to facilitate higher frequency and more comprehensive risk monitoring. - Deepening analysis of risk interactions to examine their role in aggravating systemic risk. - Expanding the macroprudential stress testing framework to include significant subsidiaries (where failure could cause major disruption and sufficient granular supervisory data is available). - Data collection priorities: - The initiative to collect cash-flow data for all significant currencies is a welcome step. - More data are needed to monitor market, funding, and contingent liquidity risks. - Cash flow data needs to be collected by significant jurisdictions to help supervisors monitor the availability and transferability of collateral. - Step up efforts on data collection on securities financing transactions and derivatives to enhance monitoring and supervision of market and contingent liquidity risks, including those associated with the flow of collateral via central clearing counterparties (CCPs) and cyber risks. ### Main recommendations on stress testing (Table 1) - Solvency Stress Testing: - Make further use of supervisory data for stress testing, enhancing analysis, and feeding back into supervision — Timing: ST - Deepen analysis of banking group risks by extending the perimeter of macroprudential stress tests to cover systemic subsidiaries of EA significant institutions and significant investment firms — Timing: MT - Continue enhancing the assessment of bank’s risk interactions across various risk categories (i.e. credit risk, market risk, basis risk, liquidity risk) — Timing: MT - Liquidity Stress Testing: - Use cash-flow data for all significant currencies and significant jurisdictions to refine liquidity monitoring and stress testing — Timing: I - Enhance collection of data on, and analysis and supervision of contingent liquidity risks associated with the cross-border flow of collateral, securities funding transactions, and cyber risks — Timing: ST - Timing legend (preserved verbatim): - I (immediate) = within one year - ST (short term) = 1–2 years - MT (medium term) = 3–5 years *EXECUTIVE SUMMARY, cr18228.* ### 2. Euro area banks represent around two-thirds of total banking system assets in ### 2. Euro area banks represent around two-thirds of total banking system assets in Europe ### Overview and key statistics - Banks are the most important financial intermediaries for households, nonfinancial corporates (NFC), and the public sector in Europe. - In 2017, total assets of the euro area banking sector, including foreign subsidiaries and branches, stood at around €35 trillion on a consolidated basis. - Since the global financial crisis, nonbanks and financial markets are playing a greater role, particularly in funding larger nonfinancial firms. ### Banking sector trends and performance - Post-crisis adjustments: - The financial crisis ended a period of strong growth in banking sector assets in many euro area countries through a reduction in business volumes and some progress in consolidation. - Large euro area banks have become more selective in their international banking activities and more focused geographically in their core markets. - Banks have shifted towards safer and less complex assets, and towards more stable funding sources, notwithstanding some legacy portfolios. - Funding and margins: - Banks have benefited from a benign funding environment with stable or improving market access in recent years, though this trend might reverse. - Euro area bank net interest margins have been resilient in recent years, supported by accommodative financial conditions which have eased funding costs helping to offset falling interest income. - The flat yield curve has enhanced banks’ reliance on commercial margins to secure interest income which might narrow due to heightened competition, including from nonbank sources. - Profitability and risks: - Although improving macroeconomic fundamentals should support headline income, performance might remain subdued for some banks. - Business risks include management of legacy assets, NPL strategies (provisioning and write-offs), fluctuations in client flows, repricing of risk in financial markets, and ongoing restructuring costs in some banks. - Investors’ perception of bank risk has lagged sovereigns although it remains more volatile than corporates. ### Supervisory stress tests in the euro area - Recent exercises: - The 2016 EBA stress test (scenario-based) found that the EU banking sector is resilient to shocks with a capital depletion of 380 bps on average. - The 2017 ECB Sensitivity Analysis of the Interest Rate Risk in the Banking Book (IRRBB) focused on shocks to interest rates on banks’ net interest income (NII) and economic value of equity (EVE); results highlighted the increasing trade-off between NII and EVE. - EBA launched its 2018 EU-wide stress test on January 2018, and results are expected to be published by November 2018. - Differences between EBA and FSAP stress tests (high level): - EBA: constrained bottom-up approach where banks project impacts subject to strict constraints defined in a common methodology; includes conduct and other operational risks; uses Through-the-Cycle (TTC) PD shifts for RWAs; assumes IFRS 9 included for banks commencing reporting in Q1 2018. - FSAP: top-down approach using in-house models; excludes conduct and other operational risks; assumes constrained balance sheets with declining credit demand but disallowed supply effects; applies Point-in-Time (PiT) PD shifts to compute RWAs; uses accounting regime valid at end-2017 for credit risk projections. - Interest rate risk: FSAP conducted granular assessment using SSM’s IRRBB templates with shocks to funding costs and lending rates calibrated at bank and portfolio level; EBA allows banks to project NII constrained by caps and floors. - Market risk: FSAP stressed fair valuation using SSM’s market risk sensitivities and viewed hedges as inefficient under stress; EBA incorporates bottom-up detailed assessment including hedges, CCR losses, and stressed CVA. ### FSAP top-down stress testing approach and caveats - Objectives and scope: - FSAP stress tests assess the euro area banking system’s ability to withstand system-wide shocks and identify macroprudential concerns complementary to the EBA bottom-up exercise. - Tests were designed to ensure banks can absorb losses without engaging in deleveraging or amplifying market and liquidity stress. - Data and methodological caveats: - FSAP stress test results are based on end-2017 supervisory data, complemented by supervisory historical data starting in September 2014; public data sources were used to fill gaps. - Matching and reconciliation of risk data from multiple sources is complex and subject to caveats. - Proxy variables were used where data availability was constrained (examples): - Duration of fixed income portfolio proxied by remaining maturity using the 2016 EBA transparency template. - PD projections for selected portfolios based on Moody’s expected default frequency (EDFs) series. - Market risk projections based on banks’ market risk sensitivities to risk factors calculated on net exposures after hedging, limiting estimation of basis risk under stress. - Stress test scenarios are calibrated using historical data to identify “tail events”; such calibration may not capture unanticipated shocks, nonlinear effects, or emerging interrelationships (examples: hard Brexit, implementation of IFRS 9, repricing risk due to basis risk and optionality, liquidity risk from contingent liabilities). ### FSAP sample of banks and coverage - Sample size and coverage: - The FSAP solvency stress test was carried out on a sample of 28 (29) large euro area banks for the solvency (liquidity) test covering broadly 65 percent (70 percent) of banking system assets. - The banks in the sample account for over 15 (16) trillion euros. - The FSAP sample for the solvency stress test covers around 85 percent of the 2018 EBA sample. - Criteria mirrored EBA’s selection: coverage of around 70 percent of euro area banking system assets and minimum size of 30 billion euros. - Due to data limitations, Credit Agricole had to be excluded from the final list of participating banks in the solvency test. - Composition and peer grouping: - The sample includes six (seven) G-SIBs of the 30 G-SIBs identified by the FSB in November 2017 for the solvency (liquidity) test. - Banks were grouped into three broad categories for benchmarking: G-SIBs, large but less complex internationally-active banks, and relatively smaller domestically-focused banks. - Table-based aggregate asset subtotals (actual data end-2017) from the sample groups: - Sub-total 8,910 (billions of euros) - Sub-total 5,207 (billions of euros) - Sub-total 2,433 (billions of euros) - Note on G-SIB capital surcharge buckets in the sample: one G-SIB in FSB designated bucket 3 (2.0 percent capital surcharge), one G-SIB in bucket 2 (1.5 percent capital surcharge), and 4 G-SIB in bucket 1 (1.0 percent capital surcharge). - Material geographies for the sample include nine home euro area countries, eleven other EU countries, and seventeen countries outside the EU (37 material geographies in total). ### FSAP stress-testing strategy and scenarios - Two-pronged approach: - Solvency stress tests: a fully-fledged scenario-based test, a range of sensitivity tests, and a targeted reverse stress test to assess capital adequacy. - The macrofinancial scenario explored domestic and global risks and included a traded risk scenario to capture business risk from tighter financial conditions, market valuation changes, and ALM vulnerabilities. - Sensitivity tests explored resilience to wider shifts in risk factors. - A reverse stress test on opaque complex assets assessed valuation shock from soft mispricing required to exhaust capital buffers of G-SIBs. - Liquidity stress tests: a wide range of tests assessing resilience to structural and idiosyncratic liquidity risks, including: - Structural liquidity analysis (LCR, NSFR, Asset Encumbrance, funding concentration). - Cash-flow tests applied under alternative scenarios simulating stress factors over five-day, one-month, and three-month maturity buckets. - Scenario characteristics (comparison with 2018 EU-wide test): - Both adverse scenarios assume materialization of systemic risks leading to a balance sheet recession in the euro area. - The FSAP scenario features long-lived shocks (U-shape) while the EBA scenario incorporates deep (V-shape) shocks. - The implied real GDP growth rate in the euro area amounts to -1.6 percent (-2.2 percent) at year t+2 under the FSAP (EU-wide) stress test by 2019. - Under the FSAP scenario the recession is more protracted with a -0.3 percent growth rate posted by 2020; the EU-wide test’s recession is more pronounced at t+2. - Both tests include a traded risk scenario linked to the macroeconomic scenario. ### Banking sector risks identified - Key vulnerabilities (unevenly distributed across countries and financial system segments): - In certain euro area countries, households, nonfinancial firms, and government bear heavy debt burdens, making them vulnerable to a tightening of rollover conditions and limiting available buffers. - Global rising asset prices and compressed spreads raise concerns given historically low market volatility and ultra-low “safe” interest rates. - Indications that credit and market risks may be being downplayed, particularly in the high-yield fixed income market and possibly equity markets. - Residential real estate prices continue to rise, particularly in some regions. - Sources of vulnerabilities noted: - Increasing exposures towards riskier segments (stretched borrowers). - Market risk from risk taking in financial markets. - Business risk from margin contraction and fluctuations in client revenue. - Operational risk from litigation. *Source: IMF staff—Technical Note prepared by Laura Valderrama and Mindaugas Leika (Monetary and Capital Markets Department, IMF), and Stefan Schmitz (external consultant); material extracted from the FSAP chapter on euro area banking system stress testing (end-2017 data).* ### 18. The euro area banking system faces a possible confluence of external and domestic ### 18. The euro area banking system faces a possible confluence of external and domestic ### Major Risks (Risk Assessment Matrix summary) - Tighter global financial conditions: - A reassessment of fundamentals amid global monetary policy normalization could trigger abrupt financial market movements. - Potential amplification by resurgence of concern over sovereign, bank, and corporate creditworthiness with adverse knock-on effects to the real economy. - Higher risk premia (including a possible decompression in term premia associated with an upside inflation surprise) and lower asset prices imply valuation losses and a reduction in the value of collateral and recoveries. - Distress in a single major institution or group of institutions could create contagious investor uncertainty given vulnerabilities and other risk factors. - Weaker global growth: - A relapse of growth in advanced economies or a significant slowdown in China and other larger emerging market countries could have a sustained negative impact on euro area exports, investment, and consumption, bringing the recent growth surge to a halt. - The impact on public and private debt sustainability may contribute to higher risk premia. - Heightened policy and geopolitical uncertainty: - Political and policy uncertainty—especially over Brexit negotiations—combined with a wide-scale retreat from cross-border integration, spurred by rising protectionism, as well as geopolitical tail risks would reduce policy collaboration, dampen the flow of capital and liquidity across countries, increase risk aversion, and could abruptly undermine the cyclical recovery. ### Solvency Stress Test — Key Elements (A) - Data and perimeter: - Stress test conducted on end-2017 data at the highest level of consolidation in the euro area. - Perimeter of consolidation of the banking group specified in CRD IV. - Insurance activities excluded; banking associates included. - Bank selection criteria: sample of banks covered in the 2018 EU-wide stress test, banks’ share in the domestic market, and banks’ role in the euro area payment system. - Data sources and inputs: - ECB confidential supervisory data post-June 2014 including EBA’s Implementing Technical Standards (ITS) covering FINREP and COREP, and data gathered via the ECB’s STE with detailed quarterly reporting on interest rate risk in the banking book and banks’ elasticities to market risk factors. - Composition of banks’ debt securities portfolio extracted from EBA’s 2016 and 2017 transparency exercise, banks’ annual reports, and Pillar 3 disclosures. - Other public sources: ECB’s MIR statistics, financial data vendors (Bloomberg, Dealogic, Haver Analytics, Moody’s KMV, Fitch), and IMF’s World Economic Outlook (WEO). - Assessment criteria (“hurdle rate”): - Hurdle rate set at the CET1 regulatory minimum of a 4.5 percent Pillar 1 requirement, a fully-phased capital conservation buffer (CCB), and a phased-in buffer for G-SIBs and O-SIIs. - This led to a CET1 hurdle rate ranging between 7.0 and 7.5 percent. - Hurdle rate also includes a 3 percent Tier 1 ratio based on the CRR leverage framework. - Other measures of soundness considered: headline income, own equity funds, and other regulatory capital definitions (Tier 1, total regulatory capital). ### Solvency Stress Test — Scenarios (B) - Horizon and baseline: - Scenario projected over a three-year horizon starting Q1 2018 extending through Q4 2020 to enhance comparability with EU-wide stress test. - Baseline scenario based on the October 2017 WEO. - Adverse scenario design and simulation: - Adverse scenario designed by the FSAP team in cooperation with the ECB to explore impact of global and domestic risks on major euro area banks. - Simulated using IMF staff’s Global Macrofinancial Model, a structural macroeconometric model disaggregated into forty national economies. - Scenario covers RAM risk factors and calibrated in line with past euro area country FSAPs: widespread growth slowdown; heightened uncertainty reflected in trade, productivity and investment shock; higher risk premia and money market rates; valuation losses on real estate, equity, and other assets. - Severity of economic contraction comparable to those in other euro area country FSAPs and the authorities’ EU-wide stress tests. - Overall scenario combines a macroeconomic recession with financial market tightening when fiscal and monetary policy buffers are largely exhausted in the euro area. - Variable coverage: - Each scenario includes 370 core macrofinancial variables encompassing domestic and international economic activity. - Core variables for each geography: growth, inflation, real investment, unemployment, policy rate, yield curve, equity prices, foreign exchange, credit growth, residential real estate prices. - Global assumptions include: world GDP growth, growth in the euro area, growth in emerging markets, fuel prices, non-fuel prices, euro area swap curve, and credit spreads by major index. ### Macroeconomic Risk Components (adverse scenario calibration) - Adverse scenario characterization: - Severe global recession in which the euro area experiences a balance sheet recession, concentrated in high-spread economies. - Overall, output contracts by 7.1 percent relative to the baseline by 2020. - Consumer price inflation falls by 3.3 percentage points. - Output loss concentrated: high-spread economies contract by 7.6 percent versus 6.9 percent in low-spread economies. - Total employment falls by 5.9 percent in the euro area. - Government debt ratio rises by 18.4 percentage points in high-spread economies and by 13.3 percentage points in low-spread economies. - Geography coverage: - 37 material geographies for the sample of large euro area banks: 9 home jurisdictions, 11 other EU countries, and 17 geographies outside the EU. - Projections include nine core euro area countries: Austria, Belgium, Finland, France, Germany, Ireland, Italy, Netherlands, Spain. - Eleven other EU countries: Bulgaria, Croatia, Czech Republic, Hungary, Luxembourg, Poland, Portugal, Romania, Slovakia, Sweden, United Kingdom. - Seventeen outside EU: Australia, Brazil, Canada, Chile, China, Colombia, Egypt, India, Japan, Mexico, New Zealand, Peru, Russia, Singapore, Switzerland, Turkey, United States. ### Market Risk Components (adverse scenario shocks) - Equity and volatility: - Real equity price falling by 20 percent in the euro area, United Kingdom, and United States over two years. - Elevated global capital market volatility widens money market spreads by 75 bps in the euro area, and by 50 bps in the United Kingdom and United States. - Government yields and term premia: - Term premium decompression and re-emergence of sovereign stress raise long-term government bond yields by: - 150 bps in high-spread euro area economies, - 75 bps in low-spread euro area economies, - 100 bps in the United Kingdom and United States. - The impact of government yield shocks under the FSAP scenario is larger than in the 2018 EU-wide stress test due to the cumulative nature of shocks in the FSAP test versus instantaneous shock approach in the EU-wide stress test. - Yield curve construction and other market instruments: - Yield curve constructed using polynomial interpolation methods: polynomial of the lowest possible degree between projected short- and long-end of the curve. - Euro area swap curve spanned based on a constrained VAR forecasting approach using exogenous path for policy rate and benchmark curves in the euro area and United States. - Projected paths for European iTraxx indices and European corporate bond yields by sector (financial, non-financial) and rating (investment grade, high-yield) generated based on exogenous path for policy rates and government spreads. - Missing data for some geographies estimated using benchmark data from comparator economies. ### Credit Risk Modelling Approach (C) - Credit risk importance and aggregate figures: - Credit risk accounts for the largest regulatory capital requirement of euro area banks. - At end-2017, the size of the largest 28 banks reached €15.0 trillion, with RWAs of €5.6 trillion, representing around 37 percent risk density. - Over 85 percent of capital is required to cover unexpected losses from credit risk. - Market risk represents under 10 percent of capital requirements. - Regulatory booking and implications: - For IRB exposures, credit risk evolves with Exposure at Default (EaD), Probability of Default (PD), and Loss Given Default (LGD). - For STA exposures, deterioration reflected in higher specific and collective allowances and higher capital requirements from credit risk downgrades. - On aggregate, two-thirds of credit risk exposures were booked under the IRB approach, with wide dispersion across banks and portfolios. - Coverage and segmentation: - 37 material geographies; COREP exposure categories included seven portfolios: central governments or central banks, institutions, corporate—SME, corporate—specialized lending, corporate—other, retail—secured by real estate, retail—other. - Other non-credit obligations assets and default fund contributions treated as corporate—other. - Separate credit risk model estimated for each bank’s material geography and regulatory portfolio. - Methodology for expected losses and capital requirements: - Loan loss impairment charges formula: EAD * PD * LGD (expressed in source as j_ti EAD LGD PD; original notation preserved in source). - Capital requirements j_ti cap calculated using conditional expected loss in Basel III’s supervisory mapping function with asset correlations and maturity adjustments prescribed in CRR/CRD IV. - PiT shifts applied to regulatory PDs for non-defaulted exposures. - Risk-weighted assets defined as j_ti EAD cap RWA = 1.25 * j_ti cap (expressed in source notation as j_ti j_ti j_ti EADcapRWA **5.12=). - Parameter pool and adjustments: - Stressed projections applied to around 10,000 initial credit risk parameters. - Pool included four key risk parameters (PD, LGD, EaD, RWA) broken down by exposure class (7 portfolios), country (12 geographies), and bank (28 banks). - Parameters assigned to obligor pool by portfolio and geography in COREP 09.02 complemented by worldwide parameters in COREP 08.02 according to banks’ modeling approach (IRB-Foundation, IRB-Advanced), asset class, and obligor grade. - Obligor grades with an implied PD=1 were excluded. - For credit exposures with geographic breakdown, defaulted exposure subtracted to estimate implied non-defaulted PDs and LGDs; adjustments made for exposures not assigned to obligor grades or pools. - Treatment of defaulted exposures and collateral: - Stressed credit risk parameters applied to non-defaulted net exposures using regulatory parameters. - No additional capital charge computed for defaulted assets to cover systematic uncertainty in realized recovery rates for these exposures on aggregate; an add-on capital surcharge on the value of the collateral was included for large defaulted portfolios. - Credit risk projections applied to projected net exposures after credit risk mitigation techniques including SME-supporting factor. - Two-step PD projection process: - Step 1: Econometric methods applied based on historical relationship between economic/financial conditions and proxies on credit risk parameters for each aggregate portfolio and country of exposure. - Step 2: Statistical analysis to adjust bank-specific projections based on portfolio proxies at the aggregate level to banks’ portfolios to capture individual risk profile of underlying exposures, drawing on post-June 2014 supervisory data. - Sovereign PD proxy: - PD proxy for exposures to central governments extracted from sovereign yields using a Merton-based approach. - FINREP returns used to back out banks’ exposures to ‘general governments’ from COREP category ‘general governments or central banks’. - A reduced-form structural model used to extract PD estimates from sovereign spreads projected in the scenario. - Using credit spreads for sovereign i linked to the scenario Si,tT, time to maturity (T-t), and assuming LGD=45 percent, the implied risk-neutral PD is backed-out as: PD = 1 - exp(-Si,tT * (T-t) / LGD) (notation preserved as in source). *International Monetary Fund — Chapter 18, euro area banking system solvency stress test (excerpts).* ### 39. PDs for institutions, corporates (including specialized lending), and retail- ### 39. PDs for institutions, corporates (including specialized lending), and retail- ### Data sources and portfolio mapping - PDs for institutions, corporates (including specialized lending), and retail-unsecured were sourced from Moody’s KMV using the one-year expected default EDF average estimate. - Categories used: the financials group, the corporate group, the construction and real estate development group, and the consumer nondurables and services group. - These categories were mapped to major COREP portfolios: institutions, corporate (including SME and specialized lending), and retail unsecured (including qualifying revolving, and other than secured lending). - For expected losses, the breakdown of loans and advances to non-financial corporations by NACE code in FINREP was mapped to Moody’s sectoral groups (example: mining and quarrying → steel and metal products group). ### Econometric treatment of default rates - To address the truncated nature of the default rate distribution, a logit transformation was applied before conducting the econometric analysis to ensure projections remain within the 0-1 bound. - Robust econometric framework included: - Time series econometric technique using Newey-West HAC-robust standard errors for heteroskedasticity and autocorrelation once regressors are stationary. - Quantile regression approach: distribution divided in quartiles (four segments); robustness checked using deciles (ten segments). - Bayesian Model Averaging (BMA) with a Normal diffuse prior distribution to address model uncertainty. ### Drivers, variables, and geographic scope - Credit risk linked to local, regional, and global drivers. Local drivers were forecasted for 37 geographies under baseline and adverse conditions. - Local variables considered: real GDP trends, a GDP-based recession indicator, inflation, unemployment, the yield curve, the spread over swap rate, credit growth, equity prices, and real estate developments. - Regional variables included: GDP growth and inflation in the euro area and emerging economies, euro swap curve developments, 3-month euro repo rates, 3-month euro Libor, European corporate spreads, EMBIG spreads, Eurostoxx 50, and sovereign debt spreads over the Bund. - Global variables included: world GDP growth, commodity prices (fuel, non-fuel), changes in 3-month U.S. dollar Libor, TED spread, changes in U.S. IG and HY corporate bond spreads, and U.S. equity prices. ### Real estate modeling and assumptions - Real house prices were projected using a long-run price and investment error-correction framework. Explanatory variables in the price equation Pt include real income (yt), stock of residential dwellings (st), and the real interest rate (rt). Investment equation uses real residential investment (ct). - Real house price indices sourced from the OECD analytical house price database; income measured by real household disposable income; interest rates measured by the long-term real interest rate. - For retail mortgages, lack of historical default time series led to using the ratio of PD projections in mortgage loans relative to corporate PDs from ECB staff’s modelling tool; this ratio was applied to FSAP team’s projected corporate PD paths to generate default rates on retail exposures secured by immovable property. - LGD projections for loans collateralized by real estate: - Assumed not to decline under the baseline scenario. - Under the adverse scenario median LGD increases by 50 percent for residential real estate and by 12 percent for commercial real estate. ### Model selection, sample periods, and robustness checks - Regressions run over two sample periods: - Full sample: quarterly data over 2005Q1 through 2016Q4. - Truncated sample: 2005Q1 through 2014Q4 to assess out-of-sample performance. - Final model selection prioritized out-of-sample forecast performance measured by root mean squared error over the period 2015Q1–2016Q4, then in-sample performance for 2005Q1–2016Q4, goodness of fit, theoretical signs of coefficients, and expert judgment benchmarked against 2008 and 2012 crises. ### Quantile regression and stress projections - Quantile regression specification for PD level of portfolio j in geography i used lagged determinants (country, regional, global factors) and the quantile λ of the conditional distribution. - Results: credit risk projections are more severe using estimated coefficients from the upper quantiles of the PD distribution. - Example: corporate exposure PDs in a low-spread country under the adverse scenario increase from 1.3 percent in 2017 to a peak of 4.0 percent in 2019 conditional on the upper quantiles; impact is 1.8 percent at the lower quantiles. - The quantile-based model identifies greater impact from sovereign stress and cliff effects from GDP contraction in the upper tail of the distribution. ### Bayesian VAR (BVAR) and spatial heterogeneity - BVAR analysis produced conditional forecasts including country, regional, and global factors; all variables endogenous except global factors considered exogenous. Model relied on a normal-diffuse prior and used the BEAR toolbox. - BVAR results suggest larger shifts to corporate PDs in the core relative to the periphery, with three clusters of PD shifts under stress across nine home euro area countries: - Elevated impact: countries 1, 3, 6, and 9. - Moderate impact: countries 2, 5, and 8. - Lower impact: countries 4 and 7. - Projections performed over a 5-year horizon but subject to wide confidence intervals in outer years. ### LGD, EAD, and capital computation methods - Stressed LGD projections informed by banks’ reported projections on defaulted exposures and ECB staff multipliers; LGD for non-defaulted portfolio backed out from COREP 09.02. - EAD projection components and drivers (for IRB portfolio): - Driven by balance sheet assumptions, structural FX risk in foreign geographies, and triggered credit lines and guarantees. - EAD change formulation included credit growth gc,t, fraction of foreign currency loans fi,c, FX depreciation ΔFXc,t, shocks to triggered credit lines and guarantees ΔLti,j, and undrawn guarantees UCLti,j. - Stressed credit conversion factors on undrawn credit lines and guarantees informed by historical off-balance sheet migration and banks’ pillar 3 disclosures. - To compute capital requirements: - ‘Hybrid’ regulatory risk parameters (PD, LGD, EAD) projected and applied to Basel III IRB formula. - RWAs computed after applying a scaling factor of 1.06 to credit RWAs and using a 1.25 multiplier to the correlation parameter of all exposures to large regulated financial institutions and to all unregulated financial institutions. - PiT shifts to PDs used to feed stressed PDs into the IRB supervisory formula; LGDs projected using a stressed recovery value on regulatory downturn LGDs. ### Standardized (STA) exposures and aggregate coverage - Estimation for STA exposures required econometric analysis to estimate: - starting value of risk parameters; - impact of the scenario on projected risk parameters; - computation of impairment flows and provisions; - impact on capital requirements based on credit rating assessments of claims. - Credit risk projections applied to around 7,000 initial credit risk parameters. - For each bank, segmentation between defaulted and non-defaulted exposures; further breakdown into the sixteen STA asset classes reported in COREP, mapped to seven IRB portfolios: government; institutions; corporate-SME; corporate-other; retail unsecured; retail secured (mortgages on immovable property); and other exposures. - Credit risk parameters included credit risk adjustments (general and specific), exposures in default (by asset class), exposure value, and risk weights. - Geographic coverage included all material geographies reported to EBA by each bank in the context of the 2016 EU-wide stress test. *Source: 39. PDs for institutions, corporates (including specialized lending), and retail-* ### 60. Two approaches were used to construct the starting point for the default rate (PD) ### cr18228 - 60. Two approaches were used to construct the starting point for the default rate (PD) ### Default-rate (PD) construction by portfolio and geography - Two approaches were used: - Stock-based approach: - Default rate = ratio of the stock of exposures in default to the amount of total exposures drawing on COREP 09.01.a. reporting. - To convert this stock-based metric into a flow-based measure, accumulation of defaults over time was projected using: - statistical analysis using the flow of new defaults and write-offs reported in COREP 09.01.b by portfolio since 2014, and - expert judgment based on individual banks’ management of problem loans (drawing on information supplied by the ECB). - IRB 1-year expected default rate approach: - Used the 1-year expected default rate reported on IRB exposures for the same portfolio and geography drawing on COREP 09.02 reporting. - This measure is forward-looking and reported as a flow; it therefore does not require additional assumptions on the flow of write-offs needed to project default rates and loan loss impairment charges. ### Coverage ratio and LGD floor - The coverage ratio for defaulted exposures was floored at the LGD rate reported for the same portfolio under the IRB approach. - Banks report the general and specific provisions computed on defaulted exposures in COREP templates; this captures eligible collateral and guarantees on defaulted claims in line with credit risk mitigation techniques. - Under the adverse scenario, a 65 percent estimated coverage ratio was applied subject to the following constraint: - 65.0 = max{multiplierLGD, exp((general_provision, specific_provision, default_in_osure, LGD, max, ic_t, ic_t, ic_t, ic_t, ic_t))} - (Note: formula text preserved as in source.) ### Forecasting flows of exposures in default and impairments - Forecasting method: - Flow of exposures in default = default rate × amount of performing exposures projected each period. - General and specific credit risk adjustments = flow of new impairments × stressed LGD by geography and portfolio. - Flow of new impairments projected using the same econometric approach applied to IRB exposures. - Balance-sheet drivers for the stock of performing exposures (IRB portfolio): - growth of the loan book, - structural FX risk in foreign exposures, - triggered credit lines and guarantees, - flow of new impairments. ### Regulatory capital requirements: three-step computation - Step 1: - Compute risk weight of exposures in default, excluding exposures at zero percent risk weight, for each portfolio using COREP 09.01.a. - Step 2: - Risk weight on the new flow of exposures in default computed as: - RWdefault_{c,i,t} = max(RWdefault_{c,i,2017}, 100%) - Step 3: - For the non-defaulted portfolio, assume one-notch downgrade of the underlying exposure under the adverse scenario. - The initial risk-weight density reported by banks for non-defaulted exposures in COREP 09.01.a was mapped to the external rating by portfolio laid out by Basel, and one-notch downgrade was added to calculate the projected risk-weight density. - This amounts to an average increase in the risk weight density for STA exposures of around 8 percent. ### Credit risk impact on profit and loss (P&L) and provisions - Expected losses: - Calculated on all exposures including on-balance sheet and off-balance sheet exposures, taking into account migration of off-balance sheet commitments to on-balance sheet. - Coverage included all asset classes for IRB and STA exposures reported in CRR. - Initial credit risk parameters = those reported for regulatory purposes in COREP templates. - Point-in-time shifts informed by models calibrated on the scenario; shifts applied to banks’ regulatory parameters to derive stressed expected losses and capital requirements. - Use of provisions: - IMF stress test allowed draw-down of provisions built for non-defaulted exposures to cover expected loan losses over the 3-year stress test horizon. - Measure of total provisions for non-defaulted exposures aggregated: - data on value adjustments and provisions on IRB exposures for obligor grades lower than 1, and - data on general and specific credit risk adjustments on STA exposures excluding exposures in default. - These provisions were used to mitigate P&L impact of credit losses over the 3-year horizon; provisions were simultaneously re-built on the flow of new exposures using scenario paths for credit risk drivers. - Cross-jurisdictional provisioning differences: - Ratio of provisions for non-defaulted loans over unimpaired loans varies between 0.4 percent and 2.5 percent across jurisdictions. - Wide disparity in provisioning practices across major euro area banks reflected in differential P&L impact under stress. - No international principles promote adoption of specific prudential tools on (general) provisions. - Interest accruals: - Interest payments assumed to accrue only on performing exposures under both baseline and adverse scenarios. - Interest revenue on performing exposures calculated on the gross carrying amount. - IMF approach is more conservative than the 2018 EU-wide stress test methodology which allows banks to project income on nonperforming exposures on a net basis. ### Effective lending rate and interest income - Effective lending rate computed on the amount of unimpaired assets at the 2017 reference date. - Interest income from loans and receivables reported in the statement of profit and loss allocated to unimpaired loans to compute the effective interest rate. - This rate was subject to rate shocks under each scenario; projected lending-rate path applied to forecast volume of unimpaired loans to generate interest income on loans and advances. ### Market risk approach: scope and methodology - Scope: - Covered all positions under fair value measurement: financial assets held for trading (HFT), financial assets designated at fair value through profit and loss (FV), trading financial assets, available-for-sale financial assets (AFS) which impact regulatory capital through OCI. - Excludes amortized cost positions held in a hedge-accounting relationship, as well as hedge accounting derivatives. - Changes in CVA and CCR are excluded. - Valuation and hedging assumptions: - All accounting categories under a full or partial fair value measurement were revalued under baseline and adverse conditions over the 3-year horizon. - Hedging instruments for interest rate risk associated to positions in the trading book and AFS portfolios are assumed to be ineffective under stress. - Traded risk losses are partially reversed as asset prices recover after a sharp correction assumed during the first year of stress. - Market risk factors and calibration: - Factors include interest rates and credit spreads for debt instruments, equity prices, exchange rates, and commodities. - Calibration of traded risk component linked to the forward-looking scenario. - Paths for market risk factors not included in the macroeconomic scenario generated using satellite models broadly aligned to the macroeconomic scenario. - For credit spreads on corporate bonds, calibration inspired by 2008 financial crisis experience. - Revenue and cost treatment: - Revenue and cost changes in investment banking business included provided they were reported under interest income. - Client revenues of trading assets projected as net interest income by applying shocks to the effective implied rate at the cut-off date using the amount reported in interest income for HFT and FV assets. - Client revenues reported under gains on financial assets and liabilities held for trading in FINREP were excluded. - Shock application: - Combined approach of instantaneous and multi-year shocks. - Instantaneous shock applied to derivative positions and equity instruments in the trading book. - Multi-year shock applied to valuation impact of debt instruments and P&L impact of net open positions in market risk factors in line with the 3-year scenario; losses absorbed the same year the shock hits. - Balance-sheet evolution constraints: - Securities portfolio notional values grow according to a formula where Pr_{j,i,t} is level of provisions for asset class j, bank i, time t; g_{i,t} is growth of interest-bearing assets for bank i; EUR_i^f, USD_i^f are fractions of portfolio denominated in EUR, USD; EUR_FXΔ, USD_FXΔ are FX shocks to EUR and USD respectively. - No portfolio rebalancing or liquidation of positions allowed throughout the stress test horizon. ### Sovereign exposures: mapping, composition, and valuation impacts - Revaluation and data sources: - Full revaluation of banks’ sovereign exposure by issuer and accounting portfolio implemented combining FINREP data with the 2016 EU-wide transparency exercise results. - Mapping between end-2017 FINREP data and 2016 EU-wide transparency exercise built to match aggregate amount of sovereign exposures at end-2017; results checked against 2017 EU-wide transparency exercise. - Market losses on sovereign exposures booked under HFT and FV were booked through profit and loss; valuation impact on AFS impacted regulatory capital through OCI. - Portfolio composition: - Average euro area bank holds around 75 percent of the sovereign portfolio in AFS; wide dispersion across banks. - Euro area banks hold around 90 percent of sovereign securities at fair value (AFS and HFT portfolios). - Some universal banks hold over 50 percent of sovereign exposures in the trading book. - Exposure magnitudes (June 2017, aggregated): - Banks hold around 9 percent of total assets in sovereign debt securities of which 30 percent are to their own sovereign, as of June 2017. - Holdings of sovereign securities as percent of total assets range between 3 percent (Finland) to 14 percent (Austria and Italy). - Own-sovereign exposure ranges between 5 percent (Finland) to just under two thirds (Ireland). - Changes over 2015–17: - Own-sovereign exposure has come down in 2015–17 and reallocated mainly outside the euro area for most banks; some banks increased sovereign holdings but reduced domestic sovereign exposure. - Valuation impact disaggregation: - Valuation impact on sovereign securities disaggregated into repricing risk and credit spread risk; impact on P&L and capital recognized over the stress test horizon. - Repricing risk calculation: - Impact depends on shock to the risk-free curve and bond duration. - Interest rate sensitivity measured by effective maturity proxied by duration approximated by residual maturity at the cut-off date. - Full valuation approach change in risk-free bond price computed as: - ΔP_{i,t} = (r_{i,t} + Δr_{i,t}) · B_{i,t} · D_i / (1 + r_{i,t}) - where D_i denotes average duration of bond i, B_{i,t} denotes carrying value of the debt security at time t, r_{i,t} captures the risk-free rate level and Δr_{i,t} the shock to the risk-free rate. *cr18228 - 60.* ### 81. The risk-free curve is proxied by the Bund yield term structure. Figure 6 shows the ### cr18228 - 81. The risk-free curve is proxied by the Bund yield term structure. Figure 6 shows the ### Risk-free curve, term premium, and sovereign stress - The risk-free curve is proxied by the Bund yield term structure. - Term premium decompression in Europe and the United States together with the absence of flight-to-quality effects contribute to steepen the Bund yield curve with the 10-year rate rising by 75 bps in 2018 relative to the starting point. - A polynomial interpolation method is applied to span the term structure between the 3-month and the 10-year rate generated in the macroeconomic scenario. - Term premium decompression together with the re-emergence of sovereign stress raises long term government bond yields further in high spread economies. - The 10-year yield on high-spread sovereign issuers reaches over 3 percent in 2018. - For countries with monetary policy space at the starting point, the expansionary policy response assumed in the adverse scenario makes the yield curve pivot downwards mitigating the pricing impact at the short-end (example: Czech Republic, United States). ### Valuation impact methodology for non-Bund holdings - The full valuation impact for other than Bund holdings depends on shocks to credit spreads and bond duration. - Using the modified duration approach, the FSAP team calculated the haircut for each fixed income instrument under each scenario by multiplying modified duration by the shocks to credit spreads: - (formula as provided)  j t j t j t i t j j t csB csr D P     - where D_j denotes average duration of bond j, B_jt denotes the carrying value of security j at time t, i_t r shows the risk-free rate, j_t cs denotes the credit spread, and j_t cs the shock to credit spreads. - Although proxied duration is typically low across banks, particularly for traded securities, the relatively larger duration of some banks’ debt holdings and the constrained monetary policy response in some countries exacerbates valuation impact in a few instances. - Duration affects valuation through two channels: (1) as the curve pivots downwards and steepens under stress, long-duration assets are hit by larger shocks; (2) the sensitivity of the portfolio to shocks increases with the duration of the portfolio. - Relatively less monetary policy space in some jurisdictions (e.g., euro area) limits the downward shift of the yield curve increasing the relative impact of term premium decompression and sovereign stress. ### Corporate exposures - Some euro area banks hold large positions in corporate debt instruments at fair value in their trading book and AFS portfolio. - Corporate exposures include debt securities issued by credit institutions, other financial corporations, and non-financial corporations. - For some banks, the size of corporate exposures at fair value is significant reaching up to 15 percent of the balance sheet. - Corporate spread shocks to major corporate yield indices were projected by rating and counterparty. Exposures split by credit rating (investment grade and high-yield), type of issuer (financial and non-financial), and geography (Europe, United States). - Projection example: Moody’s seasoned Baa corporate bond yield for U.S. corporate exposures—BAA corporate bond yields rise by 180 bps under stress. - The average duration of corporate bonds (weighted by face value) reached 3.6 years in 2017; this duration was adjusted downwards for exposures in the trading book in line with the residual maturity of the sovereign portfolio. ### Equity exposures and market-risk sensitivities - The adverse scenario assumed a broad-based sell-off in stock markets, with equity prices falling by around 20 percent in the euro area. - Loss of value on equity instruments assessed on trading book transactions and AFS portfolios; performance metrics included nominal positions, sensitivity, concentration, and holding period. - Stressed market losses benchmarked against VaR and stressed VaR (SVaR) reported in COREP. The confidence interval varied between 99.0 and 99.9 percent. - For equities held with a trading intent, fair value impact was floored using the 2018 EBA methodology: - (constraint as provided)  jshort t jlong t j t EquityEquityEq , , %20.05.1  - where the VaR scaling factor has been set to the upper bound of 1.5, and the trading position includes the fair value of equity instruments (assets) and the short positions in equity instruments (liabilities). - The FSAP team used banks’ risk sensitivities (delta, gamma, vega) collected by the SSM in the scope of the STE for market risk benchmarking. - The assessment included hedges used by banks to decrease equity sensitivities; hedges for equity risk were assumed effective under both baseline and adverse scenario. Data limitations prevented separation of cash and hedging positions. ### Derivatives and hard-to-value assets - Some large euro area banks still hold large derivative books despite overall shrinkage of wholesale activities. - In terms of market value, derivative portfolios in some banks reach over 15 percent with an upper bound of 30 percent of on-balance sheet assets. - Most derivatives are interest rate derivatives traded OTC; some banks also hold FX-derivatives. - Some derivatives are back-to-back contracts; others are position hedging with residual risk. Ability to rollover hedges under stressed conditions is untested. - Many derivatives do not have quoted prices in active markets; fair value often determined using models (e.g., Black-Scholes) with a mix of observable and unobservable inputs (volatilities, correlations, recovery rate variance). - Valuation risk in derivatives assessed via two approaches: - Use of fair-value sensitivities (greeks) to market risk parameters including standardized moves: 20 percent shift to equity prices, 100 bps increase in interest rates, 20 percent EUR/USD move, 10 percent increase in FX and equities volatility, and 20 percent increase in rates volatility. - Implementation of a reverse stress test on G-SIBs on hard-to-value assets (Level 2 and Level 3). - Hard-to-value assets include Level 3 and to some extent Level 2 assets; some euro area G-SIBs still hold material portfolios in Level 3 assets carried at fair value on unobservable inputs. - The reverse stress test assessed the size of valuation shock on Level 2 and Level 3 assets that would deplete G-SIBs’ capital buffers over SREP CET1 regulatory minimum. ### Capital requirements (standard approach) and market-risk shocks - The risk assessment included capital impact from market risk measured under the standard approach. - Most large euro area banks determine market-risk capital requirements using internal models, but the standard approach is used for some structural FX risk. - Market shocks applied to net positions included global currency pairs (GBP, JPY, CHF, CAD, USD) and local currencies in banks’ material geographies, and sharp corrections in equity positions and commodity prices for fuel and non-fuel commodities. - Euro area banks do not carry material short or long net positions under the standard approach for market risk, mitigating the impact. ### Interest Rate Risk in the Banking Book (IRRBB) — setting and implications - Euro area banks’ net interest margin (NIM) in the euro area has remained steady at around 120 bps, but this is low compared to: - 200 bps in Australia, - 230 bps in the United Kingdom, - 315 bps in the United States, - 85 bps in Japan. - Deposit rates are close to the zero lower bound, eroding the advantage of cheap deposit funding; short-end swap curve rates have become negative. - The flattening of the swap curve has constrained banks’ ability to widen the intermediation margin. The swap curve slope fell from 180 bps in 2009 to 100 bps at end-2017. - A similar pattern observed with the Bund curve with a peak-to-trough flattening of 210 bps to 100 bps in 2017. The Bund curve shows a positive slope at lower rates, posting -58 bps at the 2-year mark. - Under the baseline scenario, the 6-month Euribor is expected to increase by 100 bps by 2021 from a negative rate of 32 bps at end 2017. - While gradual rising rates are likely to support profitability in the medium term, abrupt and sharp increases in rates can erode margins further; the overall impact on regulatory capital will depend on offsetting valuation effects from assets at fair value. - Results from the 2017 ECB Sensitivity Analysis of IRRBB suggest banks are well equipped to deal with rising rates, subject to key assumptions: stability of deposits and reliability of banks’ loan pre-payment models. - EVE would be hit by a sharp rise of rates; hedges might become ineffective under stress due to basis risk arising from discrepancies between interbank rates and sight/deposit rates and mismatches between reference rates for mortgages and hedging instruments. *Source: IMF staff estimates and analysis as presented in the provided content.* ### Box 2. The 2017 ECB Sensitivity Analysis of IRRBB vs FSAP approach to IRRBB ### Box 2. The 2017 ECB Sensitivity Analysis of IRRBB vs FSAP approach to IRRBB ### Overview - Both the ECB sensitivity test of IRRBB and the FSAP assessment of IRRBB used the supervisory template collected for the short-term exercise conducted in 2017. The template shows repricing gaps for all assets and liabilities (on- and off-balance sheet) with granular maturity buckets from ‘without specified maturity’, ‘overnight’, ‘<1-month’, through ‘>20-year’. - Liabilities are split by type of contract, modeling approach, and currency. Instruments include deposits from corporates, retail deposits, deposits from other counterparties, debt instruments, and derivatives. - For the sample of banks, 55 percent of interest derivatives are net payer fixed positions to shorten banking book duration; 45 percent of interest rate derivatives are net received-fixed swaps. ### Key empirical findings and metrics - Under a hypothetical 200 bps parallel upward shift: - Average decline in EVE implies a -2.7 percent CET1 impact. - Aggregate net interest income would rise by 10.5 percent. - Under current conditions (the “year-end 2016” scenario): - Banks’ aggregate net interest income would decrease by 7.5 percent by 2020. - Impact of hedges (2016 averages): - Hedges contribute to a +1.7 percent EVE uplift on average. - Hedges have a negative average contribution to net interest income of -1.1 percent. - Sample and coverage: - ECB’s IRRBB exercise: sample of 111 significant institutions. - FSAP test: sample of 28 large banks in the euro area. - Loan and counterparty composition (stress testing sample): - Exposures to the private non-financial sector represent over two thirds of total loans. - Half of private non-financial exposures are to households (mainly mortgage loans) and half to corporates. - Around 15 percent of loans are exposures to financial corporations. - 10 percent are loans to central banks. - Over 5 percent are loans to general governments. - Funding structure (stress testing sample): - Retail funding represents over half of interest-bearing liabilities. - Deposits from financial institutions and debt instruments each contribute 20 percent to overall funding. - Average bank reliance on central bank funding is contained at 4 percent. ### Comparison: ECB sensitivity test vs FSAP approach - Shared features: - Focus: both assess impact of interest rate shocks on net interest income to evaluate earnings risk from banking book positions. - Time horizon: both used a constant portfolio structure over a 3-year horizon and drew on bank models of customer behavior. - Key differences: - Coverage: ECB = 111 institutions; FSAP = 28 banks. - Balance sheet assumption: ECB assumed constant balance sheets as of end-2016; FSAP assumed dynamic balance sheets evolving with macrofinancial conditions. - Impact assessment: - ECB used same interest rate scenarios to assess effects on capital and earnings. - FSAP used macro scenario-based assumptions to assess fair value impact from adverse movements in risk-free curves and credit spreads on debt securities (marked-to-market assets only) and separately calibrated interest rate shocks to assess P&L impact. - Calibration: - ECB: six hypothetical interest rate shocks including the “end-2016” curve, +-200 parallel moves, a steepener, a flattener, and the “end-2010” curve. - FSAP: calibrated separately the funding shock and the lending shock in line with the traded risk component of the scenario; shocks incorporated systematic risk factors and idiosyncratic shocks based on banks’ specific structures and historical behavior. ### FSAP’s four-pronged IRRBB assessment - Economic value approach: - Measures valuation effect of an interest rate shock on fair value of assets (EVE) using a market risk approach; only marked-to-market assets re-valued. - Earnings approach: - Measures impact on future net interest income over a 3-year horizon; assessed against a range of interest rate shocks calibrated at the bank/instrument level. - Scenario composition: - Includes a commercial margin and an intermediation margin. - Shock calibration includes a systematic risk factor (aligned with the scenario) and an idiosyncratic component (linked to banks’ historical rates behavior). - Application of shocks: - Shocks applied to repricing structure of assets and liabilities using the IRRBB template as of June 2017. - Impact measured separately for stock of loans/liabilities and for flow of new lending/funding instruments. ### Structure of assets and liabilities, and implications for IRRBB - Liability and funding considerations: - Impact of rate changes on funding costs depends on liability margin which varies across instruments (e.g., sight deposits vs debt securities). - Impact on interest income depends on commercial margin, which differs across loan products (e.g., mortgage vs consumer loans). - Margins reflect banks’ business models and market competition. - Behavioral uncertainty: - Market sensitivity of central bank reserves and retail funding is low, but behavioral assumptions are uncertain because these are partly motivated by the ultra-low interest rate environment and could change if rates rise. - Decoupling of rates paid to depositors from market rates increases uncertainty of deposit pricing if rates rise. ### FSAP modelling strategy and calibration - Data and mapping: - Net interest income calculated line-by-line from interest-bearing assets and liabilities using granular IRRBB repricing data mapped to banks’ asset/liability structure and implicit interest rates via FINREP templates. - Two complementary projection strategies: - Strategy 1: Projection of front-book interest rates at individual country and portfolio level. - Strategy 2: Projection of interest rates by loan product and funding instrument at the bank level. - Strategy 1 portfolio segmentation (eight segments): - Deposits from central banks; deposits from general governments; deposits from financial institutions; non-financial corporate deposits (overnight; with agreed maturity); household deposits (overnight; with agreed maturity); debt securities issued. (Deposits redeemable at notice and repurchase agreements treated as deposits with agreed maturity.) - Pricing path assumptions (selected): - Central bank and general government deposits follow changes to the monetary policy rate; policy rates assumed unchanged under stress given initial zero lower bound and deflationary environment. - Rates on deposits from financial institutions follow the LIBOR rate path assumed in the scenario. - Retail deposit pricing linked to scenario via econometric approach with thirty-six regressions for nine euro area countries and four retail categories. - Bond pricing benchmarked against euro swap curve plus bank-specific spread for new issuance. - Econometric approach: - Country-level front-book deposit rate projections used ECB MIR statistics. - Newey-West time series approach used to obtain consistent estimators with heteroskedastic and autocorrelated errors. - Regressors include autoregressive lags; macro variables (growth, inflation, credit growth); financial variables (swap rate, sovereign yield spread, policy rate, euro LIBOR, U.S. dollar LIBOR). Regressors included 1 to k lags to address endogeneity. - Empirical sensitivities: - Corporate deposit rates are more sensitive to macro conditions and money market stress. - Household deposit rates are more closely linked to the policy rate, indicating higher pass-through from policy rates to household deposit rates. ### Projected rate moves and lending/funding effects (selected projections) - Front-book deposit rate projections in an adverse scenario for a low-rated country: - Corporate deposit rates projected to rise by 140 bps by end-2021. - Household deposit rates projected to rise by 110 bps by end-2021. - The 30-bps starting-point gap produces the same end-point of 140 bps by the end of the horizon. - Lending-rate mapping and observations: - Lending-rate portfolio segmentation: loans to central banks; loans to general governments; loans to non-financial corporates (including credit institutions and other financial institutions); loans to non-financial corporations; loans to households (mortgages); other household loans (consumer loans, other). - Loans to central banks follow policy rate path; policy rate assumed unchanged in adverse scenario. - The lending rate on loans to general governments proxied by mortgage rate based on statistical analysis. - Average lending-rate levels (selected): - Average mortgage rate stood at 2.0 percent. - Average consumer loan rate stood at 5.6 percent. - Average corporate loan rate stood at 1.6 percent. - Observed divergence from risk-free curves: - The average lending rate reached 320 bps against -75 bps for the 3m Bund rate. *Source: cr18228 - Box 2. The 2017 ECB Sensitivity Analysis of IRRBB vs FSAP approach to IRRBB* ### 117. The evolution of lending rates across the euro area suggests different lending ### The evolution of lending rates across the euro area suggests different lending ### Evolution of lending rates and cross-product dispersion - Consumer loans carry a large component of credit risk premium: - Average rate on consumer loans stood at 5.6 percent in June 2017. - Mortgage rate stood at 2.0 percent in June 2017. - Corporate rate stood at 1.6 percent in June 2017. - Dispersion in lending rates (2017Q3): - Corporate rates standard deviation: 37 percent. - Mortgage rates standard deviation: 61 percent. - Consumer loans standard deviation: 150 percent. - Interpretation: Higher dispersion in consumer lending rates suggests lower competition in consumer credit relative to mortgage and corporate segments. - Lending rates versus risk-free rates: - Average lending rate across product categories reached 320 bps in 2017. - Bund rate was negative at -75 bps. - Observations: Cyclical variability of lending rates is quite limited (particularly for consumer loans), and lending rates diverge significantly from risk-free rates. ### Econometric modeling framework for lending rate forecasting - Modeling approach: - Newey-West time series approach using lagged regressors (general specification provided in the source). - Dependent variable: lending rate by country c and product j. - Regressors include: - An autoregressive variable. - Macro factors: unemployment rate, credit growth. - Financial variables: swap rate, sovereign yield spread, policy rate, euro LIBOR rate, U.S. dollar LIBOR rate. - Risk premium proxied by Moody’s EDF for corporates (corporate rate), EDF for consumer firms (consumer rate), and EDF for real estate firms (mortgage rates). - Developments in house prices (for lending rate regressions). - Regressors included 1 to k lags to address endogeneity. ### Key regression findings and behavioral insights - Product sensitivities: - Corporate lending rates are more closely linked to wholesale funding costs. - Consumer rates are driven by risk premia and show higher persistence. - Mortgage rates rise with tight credit conditions and are exposed to interbank rates due to prevalence of floating-rate mortgages linked to the 12-month EURIBOR. - Cyclicality and pass-through: - Corporate and mortgage loans are more responsive to cyclical economic conditions. - Consumer credit rates show higher persistence and greater sensitivity to market conditions, indicating higher pass-through of bank funding costs to customers—consistent with lower competition in consumer credit. - Increase of corporate rates with tightening credit conditions suggests credit rationing under stress. ### Projections under an adverse (high-spread) scenario - Lending rate projections in a high-spread country (first year of stress): - Consumer loan rates increase by 110 bps. - Mortgage rates increase by 60 bps. - Corporate rates increase by 30 bps. - Dynamics over the stress horizon: - Effect of stress is front-loaded for consumer loans (sharp first-year increase), while corporate and mortgage segments experience a more sustained gradual increase and larger increases by the end of the horizon. - Impact on net interest margins (NIM) when applying funding cost and lending rate projections to banks’ balance sheets: - On average, funding costs rise by 40 bps in 2018. - Lending rates adjust partially by 8 bps in 2018. - Banks hit by higher funding costs are on average less able than peers to pass them onto customers. - At the 90th percentile, the margin contracts by 75 bps. - Figure caption notes: At the 10th percentile, NIM contraction reaches 75 bps (boxplot description). ### Funding cost modeling: construction and data - Effective funding cost measure matches banks’ liabilities to an interest rate path and repricing structure: - Pricing of deposits from central banks and general governments follows monetary policy rates (unchanged under stress in the euro area for strategy 1). - Funding from deposit liabilities split between financial institutions (interbank deposits; interbank rates follow EURIBOR trends) and customers (households and non-financial corporations; deposit rates vary across institutions according to deposit structure and liability margin). - Debt securities (floating rate) repriced according to the swap rate; instruments include unsecured bonds (investment grade, high-yield, MTNs), secured instruments (MBS, ABS, covered bonds), and preferred securities. - Cost of fixed-rate debt split into a reference rate and a spread at issuance (coverage same as floating instruments). - Data sources and mapping: - Historical interest expense matched using Bloomberg, Fitch, and Dealogic. - Implied rates by instrument and counterparty computed using FINREP returns and mapping banks’ liability structure (F08.01.a) with banks’ interest expense by counterparty (F16.01.a). - EURIBOR time series sourced from Bloomberg. - Customer deposit historical rates computed using Fitch data. - Debt securities cost of issuance obtained from Dealogic (bonds issued at swap rate plus a spread). ### Issuance patterns, spreads, and heterogeneity - Aggregate issuance patterns (2005–2018 Q1, sample of 28 banks): - Unsecured issuance accounts for around 60 percent (mainly investment grade bonds). - Secured issuance accounts for around 40 percent, split between ABS, MBS, and covered bonds. - Issuance of preferred equity is insignificant. - Quarterly issuance fell to an average of €50 billion during crisis episodes from a pre-crisis level of €130 billion. - When bond spreads rose, gross issuance of senior unsecured bonds fell to near zero in peripheral countries and was replaced by covered bond issuance. - Cost differentials and dispersion: - Average spread of unsecured bonds: 130 bps. - Average spread of secured bonds: 60 bps. - Unsecured spread at issuance first and third quartile: 15 and 90 bps, respectively. - Secured instruments’ spread at issuance first and third quartile: 105 and 170 bps, respectively. - Wide dispersion of cost of issuance due to bank- and country-specific factors. ### Use of actual issuance spreads and econometric approaches - Rationale: - Using banks’ actual funding spreads (spreads at issuance) better captures banks’ behavioral response to stress than market proxies (e.g., 5-year CDS spreads), because banks often refrain from unsecured issuance when market spreads widen. - Econometric approaches for estimating funding cost by instrument: - Panel-based regression using fixed effects (also run at country level for robustness). - Quantile regression approach to assess drivers across percentiles of the funding cost distribution. - Basic regression specification (notation from source): - Funding cost of instrument j for bank i modeled as a function of lagged macro, financial, and regional/global factors. - Regressors include macro variables (growth, inflation, unemployment, recession indicator), financial variables (EURIBOR, swap rates, spreads over swap), and regional/global factors (growth in the EA, ted spread). - Model solved using panel fixed effects at country and euro area level. ### Empirical results on deposits, debt spreads, and sovereign-bank links - Deposit pass-through: - Pooled regression elasticity of deposit rates to changes in the ECB’s main refinancing rate: 0.2 (Annex Table 6). - Some cyclical behavior: economic expansion in the euro area and tighter money market spreads contribute to higher deposit rates. - Sovereign stress is not associated with moves in deposit rates in this analysis. - Determinants of debt issuance spreads: - Debt spreads on bank issuance mainly driven by sovereign spreads, TED spread, and bank solvency (Annex Table 7). - Sovereign-bank nexus: higher domestic sovereign stress pushes up the cost of raising market funding by reducing the sovereign rating uplift. - A shortage of global liquidity (widening TED spread) raises cost of debt issuance. - The elasticity of bank equity to debt spreads is negative and statistically significant. - Quantile regression insights: - The negative contribution of sovereign stress to the cost of debt issuance rises with sovereign stress. - At higher quantiles, the coefficient of sovereign spreads on funding costs is economically and statistically more significant than at lower quantiles—pointing to an amplification mechanism from sovereign distress to banking-sector stress through market issuance. *Source: IMF staff estimates.* ### 133. The adverse scenario is significantly more severe on the cost of debt issuance than ### cr18228 - 133. The adverse scenario is significantly more severe on the cost of debt issuance than ### Adverse scenario impact on funding costs - Deposit rates rise gradually by 15 percent, 42 percent, and 56 percent over 2018–2020 relative to the starting point. - Debt spreads increase by up to 240 percent by 2020 relative to the starting point. - Interpretation: wholesale funding is more risk sensitive than core retail deposits, hence stressed macrofinancial conditions increase the cost of debt issuance relative to deposit funding. - Figure 20: boxplots show multipliers of deposit rates (LHS) and debt spread over swap (RHS) over the stress test horizon relative to 2017; boxplots include mean (yellow dot), 25th and 75th percentiles (grey box, median indicated by change of shade), and 15th and 85th percentiles (whiskers). ### Pass-through modeling of lending rates - Approach differs from much empirical literature by using banks’ actual funding rates and banks’ asset and liability structure rather than aggregate proxies. - Notable methodological points: - Many studies estimate loan and deposit pricing using proxy rates or aggregate lending rates; this note uses banks’ actual spreads over swap. - BIS (2015) is a notable exception but uses CDS over swap as proxy rather than banks’ actual spreads over swap. - Lending rate relationship (notation preserved): - i_t^l denotes the implicit lending rate for bank i. - Equation (as presented): i_{t}^{l} = α + γ1 i_{t-1}^{l} + γ2 bankfunding_{t-1} + γ3 bankfunding_{t} + γ4 state_{t} + γ5 var + ε_{i t} - The coefficient γ3 shows the pass-through of funding costs to lending rates. - Empirical strategy: - Panel-based regressions with fixed effects and quantile regressions to assess effects across lending-rate distribution percentiles. - Results shown at aggregate and home-country levels. - Key findings on pass-through: - Autoregressive coefficient indicates relevance of banks’ back book of loans. - State of local economy affects lending rates: stronger GDP growth and wider sovereign yields contribute to higher lending rates. - Mixed evidence on pass-through of bank-specific funding costs: significant for countries 1 and 2, no evidence in other countries after controlling for financial conditions. ### Impact on net interest margin (NIM) - Data and construction: - Used ECB’s IRRBB data template to construct maturity gaps by product, maturity bucket, and currency. - Maturity buckets include: instruments without specified maturity, overnight, <1-month, 1–3-month, 3–6-month, 6–12-month, ...,1–2-year,...,>20-years. - Maturity buckets constructed separately for deposits and debt securities; debt instruments at floating rate repriced at market-based swap rate. - Debt securities within maturity bucket re-issued at projected swap rate and spread over swap, capturing systematic and idiosyncratic risk. - Net interest income shock computation (Strategy 2, notation preserved): - NII_{i t} = Σ_b (gap_{b t i} · mid_b · Δnim_{b t i}) - where gap_{b t i} is the gap of bank i in bucket b and time t, mid_b is the mid-point in bucket b, and Δnim_{b t i} is the net interest margin shock for bank i at time t. - Projected impacts (Figure 21 summary): - On average, margins contract by over 20 bps in 2018. - Banks at the bottom quartile see margins decrease by 35 bps in 2018. - Accounting for empirical estimation of pass-through effects (Strategy 2) alleviates the negative impact of stress on banks’ net interest margins compared with Strategy 1. - Figure 21 boxplots depict year-on-year projected shocks to funding costs, lending rates, and NIM (basis points, yoy) for 2018–2022 with standard boxplot statistics. ### Other profit and loss (P&L) items and fees & commissions (F&C) - Banks adapt business models toward more fee and commission-generating activities to compensate for compressed NII. - Net fees and commission income includes fiduciary activity, deposit fees, securities transfer orders and issuances, asset management services, and sales of third parties’ investment products. - Concerns: - Financial stability implications from greater reliance on F&C: competition from nonbanks and sensitivity of F&C to financial market stress may limit resilience gains from income diversification. - Reliance on F&C across sample: - Within the stress testing sample of 28 banks, the share of F&C to total assets of the average bank reached over 50 basis point in 2016. - Dispersion: universal banks rely more on F&C; first and third quartiles range between one- and two-thirds of a percentage point. - Empirical model for F&C share (notation preserved): - fc_{i t} = α + μ_{i} + μ_{c t} + μ_{t} + β1 fin + β2 macro + ε_{i t} - Model estimated using system GMM (Arellano and Bond). - Predictors include: lagged F&C over assets, contemporaneous and lagged GDP growth, inflation rate, house price growth, stock market returns, first difference of short- and long-term rate, equity volatility, government spread over swap, TED spread. - Findings: - Share of F&C increases with flattening of the yield curve and equity volatility. - Positive coefficient on short-term rate difference and negative coefficient on long-term rate difference indicate yield curve shape affects reliance on F&C. - Other P&L items projected via panel fixed-effects using Fitch time series applied to FINREP supervisory returns; main variables included: - Dividend income; - Other operating income; - Non-interest expenses including administrative expenses, depreciation, negative goodwill and other operating expenses; - Tax expenses over net profits. - One-off adjustments are treated as non-recurrent with zero P&L impact under both baseline and adverse scenarios (list includes profit and loss from non-current assets, discontinued operations, disposal groups, extraordinary profit). - Dividend payout constraints: - Floor set at 30 percent (positive profits). - Capital conservation buffer schedule for CET1. - Statistical analysis on historical dividend payments. - If profits exceeded 50 percent of income at the cut-off date, dividend payout ratio set at 50 percent. - If banks were making losses, a zero-dividend payout was assumed. ### Solvency results - Baseline scenario: - Projections imply a weighted-average 40 basis point increase in banks’ CET1 ratio at the end of the 3-year horizon relative to start, driven by retained earnings and decreased capital requirements for credit risk. - Less complex, large internationally-active banks benefit most due to improved asset quality, RWA contraction, higher lending margins in core markets, and larger loan book expansion. - Adverse scenario: - Aggregate minimum capital requirements are met, but a few banks are significantly more vulnerable. - Aggregate risk-weighted CET1 capital ratios fall by about 390 basis points in the adverse scenario, with a slightly larger impact on G-SIBs. - Less complex internationally-active banks experience smaller capital shortfalls relative to domestically-oriented banks, which incur greater-than-average hits to capital ratios. - Dispersion: coefficient of variation increases from 16 percent in 2017 to 23 percent in 2020 under stress. - Drivers of capital impact (Figure 25 summary; aggregate effects on CET1 ratio): - Weaker global growth, falling real estate prices, and tighter financial conditions: lower aggregate CET1 ratio by 3.0 percentage points relative to end-2016 due to increased credit risk and eroded collateral values. - Expansion of RWAs (heightened credit risk, valuation effects in FX exposures, projected lending paths): lower CET1 ratio by around 2.1 percentage points across banks. - Valuation losses from sharp market price movements (large trading books and AFS holdings): decrease CET1 ratio by 1.0 percentage points on average, concentrated in 2018 before partial recovery. - Interest rate risk on net income: declines CET1 ratio by around ½ percentage point on average, reflecting deposit stock, hedging practices, debt repricing structure, and partial pass-through to customers. - Pre-provision net revenue (aggregate NII, non-interest income, non-interest expenses) increases aggregate CET1 ratio by 3.1 percentage points relative to the starting point. *Source: cr18228 - IMF staff analysis (excerpt).* ### 154. The stress test results suggest that balance sheet metrics of strength and market- ### cr18228 - 154. The stress test results suggest that balance sheet metrics of strength and market- ### Stress test insights - Balance sheet metrics of strength and market-based indicators are insufficient to predict banks’ financial strength under stress. - The stress testing exercise yields additional forward-looking insights not captured by current financials; this is partly due to the use of granular supervisory data and the forward-looking nature of the exercise. - Lack of co-movement observed between various initial balance sheet metrics and capital depletion under stress: - Initial metrics considered include regulatory capital (CET1), NIM, NPLs, and CDS spreads. - While higher initial capital buffers contribute to stronger post-stressed capital ratios, alternative metrics of balance sheet strength often do not co-move with stressed outcomes. - CDS spreads do not discriminate banks under current benign market conditions but can incorporate some forward-looking information under stress in some cases. - Aggregate solvency impact highlights: - Loan impairment charges reduce aggregate CET1 ratio by about 3.0 percentage points. - Traded risk losses reduce aggregate CET1 ratio by 1 percentage point. - Rising risk-weighted assets (RWAs) reduce aggregate CET1 ratio by 2.1 percentage points. - Profitability impact: - On aggregate, banks post negative net profits of €24 billion by 2020 from a pre-stress level of €51 billion in 2017. _Source of estimates: IMF staff estimates based on ECB/SSM supervisory data and banks’ Pillar 3 disclosures and Bloomberg data._ ### Sensitivity tests (Single factor tests) - Purpose: Complement scenario-based analysis by replacing selected adverse scenario inputs with hypothetical extreme shocks and assessing sensitivity to a low-for-long interest rate environment over the 3-year horizon. - Interest rate risk: - A hypothetical 200 basis points parallel upward shift (replacing adverse scenario risk-free curve and model-based funding/lending paths) reduces the aggregate CET1 by 200 bps. - G-SIBs are relatively more impacted given the size of their trading portfolio. - Further decompression of risk premia: - A 200 bps widening on own sovereign spreads over the swap curve lowers post-stress aggregate CET1 ratio by an additional 30 bps; some less international banks are affected relatively strongly. - A widening of spreads on corporate bonds similar to that observed during 2008 leads to a further 175-bps reduction in CET1 capital; valuation losses more than offset higher earnings. G-SIBs are more severely affected given larger corporate bond exposures. - Credit rating downgrade: - A one notch-credit downgrade (on standardized exposures) erodes bank capital ratios by 60 bps. Domestically-oriented banks are impacted more severely. - Tighter LGD floor on mortgage loans: - An LGD floor of 30 percent on residential mortgages in home jurisdiction reduces aggregate CET1 ratio by around 40 bps. Effect is larger for smaller banks due to larger share of domestic mortgages. - Low-for-long: - Under the current low interest rate environment, most banks would suffer a hit to net interest income, resulting in an average CET1 reduction of 85 bps at the end of the three-year horizon. ### Hard-to-value assets (reverse sensitivity test) - Motivation: Concerns that book valuation of hard-to-value assets may be artificially high; supervisors have focused on banks’ riskiest assets and induced reductions in hard-to-value holdings. - Calibration: - Shock severity linked to instrument opacity and fair value category. - More opaque instruments (e.g., complex derivatives) hit harder than bespoke trades (e.g., reverse repos). - Level 2 assets (valued using inputs other than quoted prices) subject to less severe shocks than Level 3 assets (unobservable inputs). - Aggregate calibration specified double degree of severity for more opaque instruments and for Level 3 assets. - Only hard-to-value assets were subject to valuation shocks with no beneficial impact from fair-value adjustments in liabilities. - Results: - A combination of 10 percent valuation shock in Level 3 assets and 5 percent in Level 2 assets could deplete capital buffers of some large banks below the SREP regulatory minimum (including phased-in G-SIB capital surcharge, Pillar 2 requirement, and SRC buffer). - Estimate is conservative as stress test impact has not been adjusted for valuation reserves. ### Liquidity risk analysis and stress tests — Overview - Liquidity risk management and conditions have significantly improved in recent years; introduction of LCR and NSFR (likely as of 2019) improved regulation and reporting. - FSAP used standardized supervisory cash flow-based liquidity stress tests (multiple scenarios and horizons, e.g., 4 weeks, 3 months) rather than stressing structural LCR/NSFR directly, because average LCR requirement over calendar month is unlikely to bind in stress. - Scenarios included assumptions on central bank support, stressed market values of securities, gradual tightening of monetary conditions, changes in eligible collateral, and interest rates. - A “collateral freeze” scenario: collateral held at CCPs and available for rehypothecation remains inaccessible for five business days (e.g., cyber risk event). - Cash flow tests used supervisory contractual cash flows across maturity buckets and incorporated assumptions from solvency stress tests for consistency. ### Structural liquidity risks — Liquidity Coverage Ratio (LCR) - All banks in sample meet the 100 percent minimum LCR requirement; LCR ratios are comfortably above minimums. - Short-term overnight funding share: - Overall overnight funding is close to 22 percent; some G-SIBs and large internationally-active banks have even higher shares of short-term unsecured funding from financial corporations. - Retail funding: - Retail funding in the sample is 37 percent. - 74 percent of retail deposits are insured, bringing total share of insured deposits to 28 percent of total liabilities. - Deposit insurance implies an implied contingent liability for sovereigns or the banking system amounting to €3.6 trillion in the sample. - Smaller domestically-oriented banks: - Some have less than 10 percent of retail deposits as funding; these banks typically have higher proportion of secured long-term funding (long-term repos with CB, covered bonds, ABS) and would be vulnerable if market liquidity of secured bonds froze. - Foreign exchange funding: - U.S. dollar and British pound are most important FX exposures for sample of 29 banks. - No formal requirement to maintain 100 percent LCR for these currencies; USD and GBP LCRs are volatile and fall substantially below 100 percent for many banks. - Banks rely on U.S. dollar liquidity and central bank swap line backstops in turbulent conditions. - Collateral swaps: - Almost half of banks engage in collateral swaps, but amounts are not material relative to total LCR requirements. ### NSFR and core funding - Most banks report NSFR ratios above 100 percent. - Aggregate stable funding needs are just short of €14 billion (a surplus in banks reporting NSFR above 100 percent reaches €950 billion), driven largely by large amounts of highly liquid short-term assets and long-term repos with central banks. - Many repos with central banks exceed 1000 days of remaining maturity and account for close to EUR 500 Bn in the sample of 29 banks. - Core Funding Ratio (CFR) and funding structure: - G-SIBs rely more on short-term wholesale funding and have lower CFR compared to peers. - Unsecured retail and wholesale funding dominates banks' liabilities. - G-SIBs and internationally active banks have higher dependency on funding from financial corporations. - Smaller domestically-oriented banks display diverse funding models. ### Funding concentration and dependence on central bank funding - Overreliance on one or few funding sources increases risk; diversification and lengthening of funding reduces wholesale deposit roll-over risks albeit at higher cost. - Funding concentration has increased since 2015: - Share of ten largest funding providers rose, reaching 10 percent on average in 2017Q3. - Uptrend mainly attributed to increase in central bank funding, led by domestically-oriented banks. - G-SIBs have more diversified funding sources. - Prevalence of FX funding concentration, especially U.S. dollar, adds an additional layer of risk. *Source: IMF staff estimates based on ECB/SSM supervisory data, banks’ Pillar 3 disclosures, and Bloomberg data as cited in the source content.* ### 173. Smaller domestically-oriented banks’ dependence on CB funding reflects their ### cr18228 - 173. Smaller domestically-oriented banks’ dependence on CB funding reflects their ### Dependence on central bank (CB) funding and business models - Smaller domestically-oriented banks depend on CB funding due to their business model and challenges in obtaining cheap funding in domestic markets. - Large G-SIBs and less complex internationally-active banks have more options to diversify funding and obtain cheaper market funding from abroad. - CB funding can be used to build precautionary liquidity buffers (for example, possibly motivated by the recent bank resolution cases). - Funding from CB: - helps banks maintain higher average interest margin; - is provided under uniform policy rates while liquidity obtained depends on the quality of collateral; - may result in differing risk premiums for long-term debt issuance due to differences in sovereign risk premiums among sample countries. ### Role of CB funding in regulatory liquidity compliance - CB funding has helped banks meet regulatory liquidity requirements: - repos with private counterparties are typically short-term, while CB funding provides funding which exceeds one year (and typically goes up to five years); - this increases LCR and NSFR ratios and allows banks to manage short and long-term liquidity risks; - ability to encumber assets with central banks (especially under the full allotment auctions) provides ability to obtain cheap funding amid accommodative financial conditions. ### Impact of tightening financial conditions — channels and sensitivity - Tightening of financial conditions would affect banks unevenly, mainly via two channels: - Funding-cost channel: - Banks with higher share of central bank funding would experience relatively higher increase in funding costs. - For domestic banks, replacing the respective amount of long-term CB funding with the 2-year senior unsecured bond would lead, on average, to 10 percent higher interest expenses. - G-SIBs and internationally active banks would be much less affected. - Liquidity-ratio channel: - Long-term funding availability may affect LCR and NSFR ratios if banks cannot use less liquid assets (such as credit claims) to obtain market liquidity. - On average, the effect in volume would not be large (Box 3). - Changes in CB collateral framework and monetary policy operations may affect market liquidity of various securities and could produce wider divergence among sovereign and corporate CDS spreads, exacerbating direct funding cost effects. ### Funding-cost increases and interactions with BRRD - For some banks, funding costs will increase; for weak banks this could lead to a spread shock: - The introduction of the BRRD might reinforce the interaction between capitalization and funding costs. - An illustrative exercise (assuming replacement of CB funding with market funding via 2-year senior unsecured bonds) shows smaller, less complex domestic banks affected much more than G-SIBs or internationally active banks. - Interest rate determination in the exercise: baseline scenario forecast of 2-year swap rate plus current individual risk premiums for each bank over the respective 2-year swap rate. ### Evidence on funding concentration and CB funding (2017Q3) - Share of 10 largest funding providers to total liabilities increased, with smaller, domestic-focus banks showing greater dependency on a few funding providers (2017Q3). - CB funding is mostly long-term, with smaller domestic banks benefiting most from it. - Domestic banks would be affected disproportionally more by increase in funding costs if they were to replace CB funding with market one. ### Box 3 — Potential impact of normalization of QE and discontinuation of TLTROs - A hypothetical normalization of monetary policy (complete unwind of QE and discontinuation of TLTROs) would have multiple consequences for system-wide funding liquidity risk; direct and indirect effects on bank funding costs depend on banks’ future balance-sheet structure and market sentiment. - Key baseline magnitudes (December 2017): - QE amounted to about €2,300 billion and outstanding TLTROs to around €750 billion. - QE purchase components: - PSPP: €1,900 billion - CBPP: €240 billion - CSPP: €131 billion - ABSPP: €25 billion - Direct impact on banks’ counterbalancing capacity (CBC) under current market prices (Table 1): - PSPP: Volume €1,900 billion; Share of banks in the sample (in billion euro) 285; Assumed average haircut in CBC 4%; Impact on banks’ CBC 0.2% - CPBB/CSPP/ABSPP: Volume 396 (in billion euro); Share of banks in the sample 60; Assumed average haircut in CBC 7%; Impact on banks’ CBC 0.07% - Accounting and market-liquidity channels: - Reduction in CBC would be relatively small, dependent on haircut on swapped assets and market illiquidity of instruments accepted as collateral. - For 29 banks, prudential returns suggest haircut about 3 percent on HQLA (mostly sovereign bonds). Assuming banks account for 60 percent of the €360 billion volume proportionate to their share, reduction of their liquid CBC would amount to approximately 6 billion euro, or 0.2 percent of total CBC of the sample. - Haircut effects on less liquid instruments are larger but amounts pledged are relatively small; impacts on CBC are described as negligible in examples given. - Effects via nonbank financial institutions (NBFIs) and deposits: - Scaling down QE would reduce NBFIs’ deposits at banks. Unwinding QE would largely reduce the €1,400 billion of deposits of NBFIs (and/or the short-term deposits of households) in the sample by a similar amount. - In LCR and liquidity stress tests these deposits feature outflow rates under stress of up to 100 percent. - Banks’ deposits placed at the ECB and net exposure: - EA banks’ exposure to the ECB (current account holdings and deposit facility holdings) adds up to about €2,000 billion (banks in sample hold about 62 percent of it). - Net exposure is about €300 billion towards ECB. - Unwind of TLTROs: - TLTROs amount to roughly €750 billion in the EA; if banks in the sample participated in proportion to their share of total assets, their share would be €450 billion. - CBC would increase by the encumbered assets; EA banks have posted about €1,700 billion of collateral at the EuroSystem (after haircuts). - Least liquid assets encumbered first (credit claims, unsecured banks bonds, and ABS/RMBS) sum to about €1,000 billion — more than outstanding TLTROs of €750 billion — so liquid collateral (government bonds €340 billion and covered bonds €345 billion) can still be reported as unencumbered by banks. - Unwind would substitute CB reserves in CBC with less liquid collateral; overall aggregate impact would be small except for banks with large dependence on EuroSystem funding (smaller, less complex domestic banks). - Market-liquidity and spread-dispersion effects: - Reversal of QE effects on dispersion of spreads and/or general increase of interest rates would affect CBC via market liquidity risks. - Scenario-based effects on liquidity risk were estimated by adding potential interest rate increases and spread dispersion across EA government bonds to haircuts; effects are more important for banks that must replace larger amounts of cheap EuroSystem funding (e.g., TLTROs). - Key caveats and policy implication: - Without access to bank-level Security Holdings Statistics, estimates are rough and order-of-magnitude only. - Banks’ dependence on CB funding is uneven; some smaller, less complex domestic banks received liquidity at the expense of less liquid assets (such as credit claims). - Careful planning for phase-out of exceptional measures is important as the phase-out may affect market liquidity of securities and banks with limited ability to raise market funding. - TLTRO has effectively squeezed out some use of term funding markets, so banks highly dependent on ECB funding may face problems replacing cheap CB debt with more expensive market funding. - As monetary policy normalizes, the ECB may need to intervene in systemically important markets to maintain liquidity and price discovery. ### Counterbalancing Capacity (CBC) and Asset Encumbrance (AE) - High AE ratios among some banks hinder ability to tap unsecured funding markets; high encumbrance is common in banks that issue covered bonds to finance mortgage portfolios (notably smaller domestically-oriented banks). - Consequences for banks with high asset encumbrance: - Higher outflows from short-term market and deposit funding during idiosyncratic and systemic liquidity events; - Inability to obtain additional liquidity in the markets or central banks (central banks typically require collateral for funding). - Relative asset encumbrance ratio: - Analysis uses absolute (encumbered assets over total assets) and relative (encumbered assets over total assets which might be encumbered — i.e., only liquid assets are counted in denominator). - The relative encumbrance ratio average in the sample is 65 percent. - Average ability to encumber ratio across 29 banks is 35 percent (i.e. 100–65 percent). - Relative ratio reflects available liquidity and how much assets a bank can quickly encumber in a case of liquidity stress; averages do not reflect individual bank risks. - Contingent liquidity risks: - Based on bottom-up sensitivity analysis regularly reported by banks, €738 billion of additional collateral would need to be posted were collateral value to decline by 30 percent. - Scenarios are based on gross numbers and do not reflect value of collateral on the receiving leg (thus might overestimate total effect). - This represents a 23 percent increase in aggregate level of encumbrance and some of the 29 banks would face challenges in meeting all margin call requirements. - CBC heterogeneity and implications: - CBC is the first line of defense in liquidity shocks; individual banks are heterogeneous in amount and quality of CBC. - Dispersion is observed among less complex, internationally-active and smaller, domestic-oriented banks. - Some smaller domestically-oriented banks have higher share of long-term market-based or CB-based funding and thus may need lower CBC. - Banks with higher CB AE ratios tend to have higher quality CBC because the ratio measures quality of remaining unencumbered liquid assets. *IMF staff analysis as presented in cr18228 - 173.* ### Box 4. Contingent Liquidity Risks ### Box 4. Contingent Liquidity Risks ### Overview of contingent liquidity risks - Contingent liquidity risks arise from financial contracts, such as repos (or securities financing transactions (SFT)) and derivatives. - These risks are contractual but contingent on market events (e.g., change in value of securities or a derivative being in or out of money). - Importance increased due to migration of OTC SFTs and derivatives into centralized clearinghouses and bilateral contract margining requirements. - Typical mechanics: a market-wide shock can reduce the value of collateral a bank posted while the bank may simultaneously receive collateral on opposite transactions (reverse repos). Collateral is typically cash or sovereign bonds, sometimes less liquid assets (e.g., equities). - Netting of posted and received collateral is crucial: if a bank is market-neutral it will both post and receive collateral and liquidity needs are netted—especially when contracts are cleared via CCPs. ### Role of CCPs and netting - Net effect of matched positions mitigates contingent liquidity risks; central clearing (CCPs) enhances netting and reduces gross collateral demands. - If a bank must post more collateral than it receives, its AE ratio goes up, which dries out CBC and may lead to idiosyncratic shocks (e.g., inability to roll over unsecured wholesale funding). ### Asset encumbrance, AE and bank vulnerability (select findings) - Banks with low RWA density and high AE may be subject to heightened liquidity risks, particularly if losses in mortgage segments sharply reduce CAR when refinancing needs occur (e.g., redeeming own debt securities). - Banks with high LCR ratios tend to have higher AE ratios, indicating reliance on liquidity transformation and support from central banks. ### Cash-Flow Liquidity Stress Tests (CFLST): objectives and scope - Objective: analyse liquidity risk exposure and risk bearing capacity of a sample of 29 banks in the EA. - CFLSTs estimate potential liquidity needs of individual banks and the banking system under a baseline and multiple stress scenarios, revealing liquidity risk tolerance and common exposures (e.g., reliance on unsecured short-term funding, holdings of similar less liquid assets in CBC). - CFLSTs do not consider potential redistribution of liquidity within the banking system (e.g., migration of deposits from capital-short banks to well-capitalized banks). ### Method, data and indicators - CFLSTs transform reported cash-flow data into stressed cash-flows and security flows using scenario-dependent stress factors. - Two key indicators: - Net-funding gap (NFG) = difference between cash-inflows and cash-outflows in each time bucket; cumulated net-funding gap (CNFG) = sum across buckets. - Counterbalancing capacity (CBC) = sum of cash inflows banks can generate under stress at reasonable prices in the respective bucket after accounting for securities flows. CCBC = sum of counterbalancing capacities across time buckets and the current one. - Data: Additional Maturity Ladder (C66.00) from 4Q 2014 to 3Q 2017 (Short-Term Exercise). - SSM data quality assurance: three-step process involving NCAs, SSM statistics department, and JSTs; IMF performed additional checks comparing cash flow data with balance sheet and other templates. ### Key quantitative findings from CFLST - Contractual liquidity risk exposure is high: contractual outflows within the first 4 weeks amount to about 31 percent of TA (weighted average; excluding open maturity and overnight retail deposits (23 percent of TA) and open maturity and overnight corporate deposits (7.5 percent of TA)). - Contractual inflows amount to about 21 percent of TA (excluding inflows from central bank deposits (5.9 percent of TA) due to reporting conventions). - The cumulated net funding gap over the first 4 weeks reaches about 10 percent of TA or €1,700 billion. - Main drivers of the net outflows: - Outflows from deposits of financial institutions: -4 percent of TA (net). - Other deposit outflows: -2.6 percent of TA (net). - Repos collateralized with 0 percent risk-weight bonds: -1.4 percent of TA (net). - Gross encumbrance due to repos across all asset categories: about 14 percent of TA; gross reverse repos: about 10 percent of TA. - From aggregated reported buckets: total contractual cash outflows across time buckets reported as 61.28 percent (table total); total cash inflows across time buckets reported as 26.95 percent (table total). ### Counterbalancing Capacity (CBC): size and composition - System-wide CBC in the first month fully covers the system-wide cash flow gap. Total sum of assets in the CBC amounts to 20 percent of TA, higher than the net funding gap in the unstressed reported data (about 10 percent of TA over the first four weeks). Distribution is uneven and some banks face shortfalls under stress. - CBC composition (Weighted average across all banks, in percent of TA, 2017 Q3): - HQLA I / Marketable liquid assets: - Cash: 0.3 - Central bank deposits: 7.3 - 0 percent risk-weight securities: 5.8 - HQLA II: - 20 percent risk-weight securities: 0.2 - Covered bonds: 0.8 - Corporate bonds: 0.2 - RMBS: 0.5 - Non-HQLA / Non-Marketable liquid assets: - Other CB eligible assets (i.e. Credit Claims): 3.4 - Non-CB eligible equity: 0.7 - Other non-CB eligible assets: 0.8 - Most securities in CBC are low to very low credit risk (CQS1). Banks active in FX funding markets accumulated U.S. sovereign and U.S. government-sponsored enterprise papers, also CQS1. - Euro area and U.S. sovereign securities are the most prevalent debt assets in CBC: five countries (Italy, Spain, Germany, United States, and France) cover roughly 65 percent of sovereign exposures in the sample. Average remaining maturity of these sovereign securities is close to two years. ### Scenarios and parameter uncertainty - Parameter uncertainty addressed by running a broad set of embedded scenarios of increasing severity. - Scenario set: 20 embedded scenarios for 4-week and 3-month horizons and five scenarios for a 5-day horizon: baseline plus 4 stress scenarios with increasing severity (mild market stress, medium market stress, severe market stress, combined severe market-wide and idiosyncratic stress). - Each stress scenario combined with four approaches to CBC: - (i) full CBC without haircuts: fully endogenous central bank liquidity supply if banks have unencumbered eligible collateral; - (ii) full CBC with bank specific haircuts; - (iii) marketable CBC: disregard non-marketable components of CBC (e.g., credit claims, committed lines); - (iv) liquid CBC: bank specific haircuts and bank specific market price effects derived from solvency stress test adverse scenario for assets liquid in private markets. - Calibration of haircuts under liquid CBC draws on asset prices under the adverse scenario of solvency stress tests to ensure consistency. ### Special five-day scenario: cyber-risk and rehypothecation - Five-day cyber-risk scenario assumed unavailability of collateral used for rehypothecation. Rehypothecation means a financial institution uses collateral received from another institution to cover its own liquidity needs; unencumbered collateral available for rehypothecation is included in CBC for the life of the loan contract. - Scenario assumed a severe liquidity event lasting five working days with daily simulation of contractual and behavioral cash flows. Choice of five days based on assumption that access to collateral would be restored at the end of this horizon. - Data constraints prevented separation of bilateral and tri-party repos (bilateral repos are less susceptible to cyber events as they may be managed bilaterally and not via CCPs or a central depository). *Source: Box 4. Contingent Liquidity Risks, cr18228 (IMF).* ### 196. Scenario calibration builds on event studies of system-wide and idiosyncratic ### Scenario calibration builds on event studies of system-wide and idiosyncratic ### Evidence and benchmark calibrations - Scenario calibration builds on event studies of system-wide and idiosyncratic liquidity stress events and is broadly consistent with the literature. - Historical retail deposit outflows cited: - Banesto (ES, 1994): 11 percent within one week. - Indymac (USA, June 2008): 8 percent. - IndyMac (USA, June 2008) [alternate figure in text]: 7.5 percent. - Washington Mutual (USA, September 2008): 8.5 percent in 10 days. - DSB Bank (NL, 2009): 30 percent in 12 days (Schmieder et al. 2012, Table 3). - Example severe scenario (comparable to the Lehman crisis) applied in literature: - Retail term deposits outflow: 10 percent over the 30-day horizon. - Demand deposits outflow: 20 percent. - Wholesale and other run-off benchmarks from literature: - Unsecured short-term wholesale funding run-off rates: 100 percent. - Secured wholesale funding outflow rate: 20 percent. - EBA Severe Market Scenario run-off rates: - Retail deposits: 5 percent. - NFC deposits: 10 percent. - Nonbank financial institutions: 20 percent. - Financial institutions: 100 percent. - Government/public entities: 0 percent. - Hałaj, Laliotis (2017, Severely adverse scenario) run-off rates: - Stable deposits: 10 percent. - Non-stable deposits: 20 percent. - Net unsecured interbank funding: 100 percent. - Net secured interbank funding: 50 percent. - Other wholesale funding (except ABS): 100 percent (ABS 50 percent). ### Run-off and inflow parameterization - Run-off rates higher for unsecured than for secured wholesale funding, and higher for non-insured deposits than for insured ones. - Inflow parameters: - In principle 100 percent of contractual inflows, except inflows from loans to retail and corporate customers: 0 percent. - Rationale: CFLST assumes banks continue business as normal to analyze banks’ ability to cope with liquidity stress while maintaining lending to the real economy. ### Haircuts, CBC composition, and methods - Haircuts capture: - Haircuts applied by the repo counterparty. - Market price effect (value upon which the haircut is applied). - Four approaches to counterbalancing capacity (CBC) in the analysis: - Full CBC approach without haircuts: no haircuts; full CBC as reported by banks included. - Full CBC approach with haircuts: bank specific haircuts applied; full CBC included. - Marketable CBC: bank specific haircuts applied; haircuts for “Other central bank eligible assets” and for “Undrawn committed lines” provided by other banks set to 100 percent. - Liquid CBC: same as Marketable CBC, plus ECB haircuts applied on top of market price changes derived from the solvency scenario. - Bank-specific haircuts: - CFLST refined to apply bank specific haircuts rather than same haircuts for all banks because CBC composition and haircuts differ across banks. - Assumption: haircuts applied to collateral already deposited with the Eurosystem equal haircuts for remaining unencumbered assets (caveat: may not fully reflect haircuts for assets available for encumbrance). - ECB haircuts (averages across banks, Table 8): - Cash: 0.0 (Average haircut across Banks). - CB reserves: 0.0. - 0percent RW securities: 4.2 (Standard Deviation 3.6; Max-Min 12.1). - 20percent RW securities: 3.3 (Standard Deviation 2.5; Max-Min 9.5). - Covered bonds: 7.1 (Standard Deviation 3.9; Max-Min 14.5). - Corporate bonds (NFC): 7.1 (Standard Deviation 3.9; Max-Min 14.5). - Residential Real Estate Mortgage Backed Securities: 5.6 (Standard Deviation 4.7; Max-Min 15.5). - Other central bank eligible assets (credit claims): 26.1 (Standard Deviation 13.0; Max-Min 47.5). - Equities: 25.0–75.0. - Other non-central bank eligible assets: 75.0. - Undrawn committed lines provided to the bank: 50.0. - Unencumbered collateral haircuts and market impact: - Reported average haircuts for sovereign bonds in unencumbered assets template around 10 percent (slightly more conservative than average ECB haircuts). - Equity instruments have haircut close to 100 percent (assumes almost illiquid within CBC). - For impact of market prices on CBC, haircuts on unencumbered sovereign and corporate securities were calculated under the adverse scenario of the solvency stress test, which produced two components: changes in risk free-rates (yields of German sovereign bonds) and risk premiums over the German sovereign reference rate. - Market repo haircuts by collateral type and duration (Figure 34, in percent): - Government securities: 1.6 (1 m), 1.9 (3 m), 2.1 (1 y), 2.4 (>1 y). - Public agencies / sub-national governments: 2.0, 2.3, 2.6, 2.9. - Supranational agencies: 1.8, 2.2, 2.4, 2.7. - Corporate bonds (financial): 4.3, 5.1, 5.6, 6.4. - Covered bonds: 2.6, 3.1, 3.4, 3.9. - RMBS/CMBS: 3.8, 4.6, 5.0, 5.7. - Other asset backed: 3.4, 4.0, 4.4, 5.0. - CDO, CLN, CLO: 3.6, 4.3, 4.8, 5.4. - Convertible bonds: 6.0, 7.1, 7.8, 8.9. - Equity: 4.8, 5.7, 6.2, 7.1. - Other: 3.4, 4.1, 4.5, 5.1. - Example haircuts by sovereign (LCH Clearnet margin collateral haircuts, by term of security; selected entries): - Austria: 5.5 (<=1yr), 6.3 (<=3yrs), 8.0 (>3yrs <=7yrs), 8.3 (>7yrs <=11yrs), 11.5 (>11yrs <=30yrs), 13.6 (>30yrs). - France: 5.8, 6.3, 7.3, 8.1, 11.8, 13.9. - Germany: 5.5, 6.1, 7.3, 8.0, 12.3, 12.8. - Italy: 8.0, 10.5, 14.1, 15.9, 19.8, 21.6. - CBC composition: Type of security totals and haircuts (sources cited): - Securities issued by general governments: Total amount 706 (billions of euro); CB eligible amount 634 (billions of euro); Haircuts 10 (in percent). - Securities issued by corporations: Total amount 101; CB eligible amount 85; Haircuts 16. - Asset-backed securities: Total amount 32; CB eligible amount 25; Haircuts 22. - Equity instruments: Total amount 8; CB eligible amount 0.4; Haircuts 96. ### Results: Cash Flow-based Liquidity Stress Tests — 4-Week time horizon - Interaction of scenario severity and central bank reaction reported in results matrix (Tables 9–11). - Aggregate and bank-level outcomes: - Even the mild market scenario leads to a reduction of the aggregate CBC of 5 percent of total assets (Table 9). - Under the most severe scenario, impact amounts to 16 percent of total assets or 80 percent of the initial CBC. - All banks have positive CCBC throughout the stress horizon under the full CBC approaches without haircuts (HC), except in the severe Idiosyncratic scenario where one bank features a negative CCBC after stress. - Disregarding non-marketable components of the CBC leads to a maximum of 5 banks with negative CCBC (Table 10) with an average shortfall of 4 percent of total assets of these banks (Table 11). - Under the “Liquid CBC” the severity of stress reaches 16 percent of total assets; 6 banks feature a negative CCBC after stress with a shortfall of 3 percent of their total assets. - From a system-wide perspective, the aggregate shortfall of 0.6 percent of the total assets of the entire sample is small. - Table 9. Average Scenario Impact in Percent of TA in the 4-Week Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline 0; Mild Market -5; Medium Market -8; Severe Market -12; Severe M/Idiosyncratic -14. - Full CBC w HCs: Baseline 1; Mild Market -4; Medium Market -7; Severe Market -11; Severe M/Idiosyncratic -13. - Marketable CBC: Baseline 1; Mild Market -7; Medium Market -10; Severe Market -14; Severe M/Idiosyncratic -16. - Liquid CBC: Baseline 1; Mild Market -7; Medium Market -10; Severe Market -14; Severe M/Idiosyncratic -16. - Table 10. Number of Banks with Neg. CCBC in the 4-Week Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market 0; Severe M/Idiosyncratic 0. - Full CBC w HCs: 0; 0; 0; 0; 1. - Marketable CBC: 0; 0; 0; 4; 5. - Liquid CBC: 0; 0; 0; 4; 6. - Table 11. Average CCBC (in percent of TA) of Banks Neg. CCBC in the 4-Week Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market 0; Severe M/Idiosyncratic 0. - Full CBC w HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market 0; Severe M/Idiosyncratic -3. - Marketable CBC: Baseline 0; Mild Market 0; Medium Market 0; Severe Market -2; Severe M/Idiosyncratic -4. - Liquid CBC: Baseline 0; Mild Market 0; Medium Market 0; Severe Market -2; Severe M/Idiosyncratic -3. - Contributions and magnitudes: - Total contribution to decline in CBC if only marketable collateral is used: €136 billion, or 8 percent of initial CBC. - In the most severe case, total decline in CBC (after market price haircuts, bank specific haircuts and assuming liquid collateral only) is close to 20 percent of the initial CBC. - To cover outflows, banks would use cash and deposits at the central bank; repos and reverse repos form a sizable portion of inflows and largely compensate for repo outflows. ### Results: Cash Flow-based Liquidity Stress Tests — 3-Month time horizon - Aggregate resilience but some banks approach critical positions under medium and severe scenarios. - Under the most severe scenario: - Impact amounts to 18 percent of total assets or 90 percent of the initial CBC (Table 12). - All banks have positive CCBC throughout the stress horizon under the full CBC approach without HCs in all but the most severe scenario, in which five banks display a negative CCBC of 1 percent of their total assets (Table 13 and Table 14). - Disregarding non-marketable components of the CBC leads to a maximum of 11 banks with negative CCBC with an average shortfall of 3 percent of their total assets. - From a system-wide perspective, the aggregate shortfall is 1.3 percent of the total assets of the entire sample. - Table 12. Average Scenario Impact in Percent of TA in the 3-Month Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline -1; Mild Market -6; Medium Market -10; Severe Market -14; Severe M/Idiosyncratic -16. - Full CBC w HCs: Baseline 0; Mild Market -6; Medium Market -9; Severe Market -14; Severe M/Idiosyncratic -16. - Marketable CBC: Baseline 0; Mild Market -8; Medium Market -12; Severe Market -16; Severe M/Idiosyncratic -18. - Liquid CBC: Baseline 0; Mild Market -8; Medium Market -12; Severe Market -16; Severe M/Idiosyncratic -18. - Table 13. Number of Banks with Neg. CCBC in the 3-Month Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market 0; Severe M/Idiosyncratic 5. - Full CBC w HCs: 0; 0; 0; 2; 5. - Marketable CBC: 0; 0; 1; 4; 11. - Liquid CBC: 0; 0; 2; 7; 11. - Table 14. Average CCBC (in percent of TA) of Banks Neg. CCBC in the 3-Month Time Horizon Under Various Scenarios: - Full CBC w/o HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market 0; Severe M/Idiosyncratic -1. - Full CBC w HCs: Baseline 0; Mild Market 0; Medium Market 0; Severe Market -1; Severe M/Idiosyncratic -2. - Marketable CBC: Baseline 0; Mild Market 0; Medium Market -1; Severe Market -4; Severe M/Idiosyncratic -3. - Liquid CBC: Baseline 0; Mild Market 0; Medium Market -1; Severe Market -2; Severe M/Idiosyncratic -3. ### Key drivers of stress and vulnerabilities - Major drivers of liquidity stress: - Net unsecured deposits of financial institutions. - Net repos. - Retail, corporate, and other deposit outflows, as well as scenario-based haircuts on the CBC, are substantial only in the most adverse case. - Bank types most vulnerable: - G-SIBs (Global Systemically Important Banks): more dependent on short-term wholesale funding from financial corporates and thus experience higher funding outflows. - Domestic banks: some less complex, domestically oriented banks tend to have lower CBC buffers and are more affected by changes in CBC liquidity assumptions. *Source: cr18228 - 196. Scenario calibration builds on event studies of system-wide and idiosyncratic (PDF chapter/section).* ### 209. Cash flow patterns under the 3-month scenario are similar to the 4-weeks scenario, ### cr18228 - 209. Cash flow patterns under the 3-month scenario are similar to the 4-weeks scenario ### 3-Month Liquidity Stress: key results - Cash flow patterns under the 3-month scenario are similar to the 4-weeks scenario because many short-term cash flows mature within the first few weeks. - Aggregate CBC drops from €3.2 trillion to just €45 billion under the 3-month severe/idiosyncratic, liquid CBC scenario. - Remaining aggregate CBC under the same four weeks scenario drops to €408 billion (for comparison: 3-month scenario = €45 billion). - The sharp decline in aggregate CBC is driven by other inflows and outflows, and a slightly higher proportion of repos and reverse repos maturing. - A few more banks, especially internationally-active ones, feature a negative CBC under the severe/idiosyncratic, liquid CBC three-month scenario compared with the same four weeks scenario. - Despite large aggregate depletion, the majority of banks still maintain a liquidity surplus above zero; however, a few less complex, internationally active banks would see their CBC negative. ### Rehypothecation, collateral access, and 5-day collateral freeze scenario - G-SIBs active in repos and derivatives markets are more dependent on rehypothecation for liquidity management: share of collateral received and available for rehypothecation can be a significant part of CBC in some G-SIBs and internationally active banks. - Smaller, less complex domestic banks rarely engage in rehypothecation transactions. - Cyber-risk event scenario assumption: banks cannot access collateral located in CCPs, CSDs, etc., e.g., due to a cyber-attack on market infrastructure. Collateral used for bilateral repo and derivatives (OTC, not cleared through CCP) may not be affected. - Under the five-day collateral freeze scenario: - In the baseline scenario, no bank’s liquidity buffers were depleted. - In the adverse/severe scenario, a few banks failed the test. - Collateral unavailability in the severe scenario within the first three days does not lead to negative CBC for any banks; a few banks failed over the remaining two days. - Total impact on CBC from this event is 9 percent decline; the distribution of impact is uneven, affecting G-SIBs and internationally-active banks more. - Difference vs. four weeks severe, liquid CBC scenario: a few banks fail the test much sooner under the five-day cyber scenario (no bank fails within the first week in the four weeks scenario). - Limitations: the test does not assume cascading liquidity effects on positions of other market participants (e.g., consequences if a G-SIB becomes illiquid and cannot provide services). Liquidity risks from rehypothecation need further analysis and monitoring, ideally using granular institution-level data about contingent liquidity flows from derivatives and securities financing transactions. ### System resilience and heterogeneity - The general resilience of the system to liquidity shocks has substantially increased in recent years amid a generally accommodative financing environment. - Extraordinary Eurosystem operations have changed the composition of banks’ liabilities and the structure of their CBC. - Banks differ notably in funding structures, asset encumbrance, and amount/composition of CBC. - Distributional results: - The distribution of CCBC after stress shifts left with increasing scenario severity; most banks remain liquid even in the most severe scenario (combined severe market and idiosyncratic scenario over 3 months with “marketable CBC”). - Banks with negative CCBC after stress are heterogeneous in causes: low initial CBC, large shares of credit claims in initial CBC, high outflows from committed lines to customers, or high outflows from deposits of financial institutions or other deposits. - Financial tightening impacts banks unevenly: - Banks more dependent on central bank refinancing operations (e.g., long-term refinancing operations), and those that used lower quality collateral to obtain liquidity, are most at risk of higher funding costs when reverting to market funding. ### Policy recommendations (liquidity and funding) - Banks should proactively strengthen balance sheets to ensure reasonable funding costs as reliance on private markets increases: - Lengthen funding tenors. - Increase liquidity buffers. - Increase CBC where needed. - Lengthen and stagger tenors of deposits from financial institutions. - Improve risk-sensitivity of pricing of committed lines to customers. - Supervisors should focus attention on banks that feature a negative CCBC after stress. - Prepare for tightening of financial conditions: - Compensate changes in monetary conditions and increase in policy rates to maintain profitability. - Options to minimize effect of increased funding cost: pass-through of increased costs to interest rates on new loans; increase in fees and commissions; longer maturity transformation (substitution of short-term investments into sovereign/low yield securities with longer-term assets). - Longer maturity transformation requires increase in long-term stable funding; feasible for banks that cleaned up balance sheets, adopted cost-efficient business models, and can issue equity in markets. ### Exploratory study: Solvency-Liquidity Integration — approach and high-level findings - Motivation: quantify amplification and feedback mechanisms between solvency, liquidity, and funding costs; capture endogenous effects of changes in CET1 ratios on funding costs over stress horizon and liability structure changes under regulatory constraints (LCR, NSFR). - Threefold approach: - Funding cost feedback loop: quantify feedback loop between funding costs and capital ratios under adverse scenario. - Credit rating analysis: project implied credit rating or CDS spread using a Merton-based approach based on solvency stress test inputs. - Backbook effect: estimate required pass-through rates on flows of matured loans (assuming constant balance sheet) to absorb higher funding costs. Module 1 — Feedback loop analysis: methodology and results - Iterative projection process at each quarterly time step between 2018Q1 and 2020Q4: - Stage 1: Initial projection of effective funding costs by mapping liability structure to interest rate path for each instrument; interest paid on each liability conditional on bank’s initial capital ratio. - Stage 2: Initial projection of effective interest income using econometric pass-through at bank level; effective interest received depends on loan book composition, market reference rate, and bank commercial margin linked to scenario; lending rate dependent on bank’s initial capital position. - Stage 3: Initial projection of bank CET1 capital combining scenario and initial conditions to project losses, revenues, costs, and balance sheet items. - Stage 4: Iterative process where projected CET1 informs revised effective funding costs and effective interest income; revised net interest income combined with projections to update bank capital; iterate until convergence. - Effective bank deposit rates model: deposit rates sensitive to bank solvency. Econometric finding: - A 100-bps increase in regulatory Tier 1 capital is linked to a 20 bps decrease in the effective rate on customer deposit. - Relationship is convex (positive coefficient on squared term). - Effective lending rates model: pass-through of funding costs to lending rates is conditional on bank characteristics. - Econometric result: pass-through coefficient estimated at 0.65 and statistically significant at the 99 percent confidence level once bank-specific variables are included. - Funding costs are passed through more quickly to lending rates when capital buffers are lower (weaker banks tend to increase pass-through more completely). - Iterations and multipliers: - Convergence achieved after three iterations. - Multipliers (ratio of stressed funding cost under adverse scenario to starting position) for deposit rates — median across banks: - Iteration 1: 2018 = 1.14, 2019 = 1.38, 2020 = 1.58. - Iteration 2: 2018 = 1.37, 2019 = 1.83, 2020 = 2.21. - Iteration 3: 2018 = 1.37, 2019 = 1.84, 2020 = 2.22 (and remain stable in successive iterations). - Impact on capital: - Accounting for feedback loops in funding costs adds an average 30 bps to capital depletion. - At the 90th percentile the impact increases to 50 bps. - Weaker banks are particularly hit, amplifying an adverse feedback loop. *Source: IMF staff calculations, cr18228 - 209. Cash flow patterns under the 3-month scenario are similar to the 4-weeks scenario, (PDF chapter).* ### 230. Credit ratings provide an indication of funding costs and funding availability which ### 230. Credit ratings provide an indication of funding costs and funding availability which ### Credit-rating methodology and purpose - Credit ratings provide an indication of funding costs and funding availability which are complementary relative to regulatory capital ratios. - The analysis quantifies how much the credit rating of the bank will change under the stressed operating environment evaluated in the solvency stress test. - The analysis was based on the Bloomberg’s DRSK module based on a Merton-like approach. This function calculates the implied default probability, implied rating, and implied CDS based on current bank data and is overridden with the paths projected in the solvency stress test to generate implied shifts to credit ratings. - The DRSK model is a balance sheet focused measure of credit risk based on the structure model proposed by Merton: - A bank is viewed as solvent if the value of the assets is larger than the value of its liabilities. - Equity is viewed as a call option on total assets where the strike price equals liabilities. - The value of bank assets is inferred from the equity value using a Black Scholes option pricing approach. - Trading liabilities (repo, short sales, and derivatives) are important measures of credit risk alongside short-term debt; customer deposits receive a 50 percent haircut. ### Model inputs and projections - Key inputs (from the worst period in the FSAP solvency stress test) — eight inputs in total: - Market category (3 inputs): share price, market capitalization, and share price volatility. - Financials category (5 inputs): effective short-term debt, effective long-term debt, loan loss reserves, nonperforming loans, and net income. - Individual bank market variables are projected in line with the scenario: - A simple model estimates betas of individual bank share prices combined with the path of the equity share price for major indices in the euro area calibrated in the scenario. - The correction in equity prices is accompanied by a sudden drop in market cap and an increase in volatility. - Financial variables were extracted from the solvency results; financials are projected as of end-2020. - Loan loss reserves were calculated using a stressed coverage ratio on defaulted exposures. - Nonperforming loans were projected for IRB and STA exposures. - Net income was estimated from P&L projections. - The Merton-based model relies on the distance-to-default leverage parameter incorporating firm value to debt adjusted for market volatility and expected return on assets (extracted from equity prices), but inclusion of additional financial variables can improve model performance. ### Credit-rating migration results under stress - Results suggest significant non-linear amplification effects from the combination of: deterioration in banks’ financials, liquidity stress, and a market price correction. - Specific impacts on implied PDs for listed banks (averages reported): - Deteriorating economic conditions under the adverse scenario: implied PD increases by [80] bps. - Additional liquidity shock (long-dated debt dries up and banks forced to increase reliance on short-term maturities by 20 percent): implied PD increases by a further [40] bps. - Share price decline envisaged in the scenario (major indices declining by 20 percent): additional impact of [60] bps increase in default risk. - Mapping to Bloomberg Credit Risk Scale: the combination of solvency shocks, liquidity risk, and equity correction lowers banks’ credit rating by [four] notch-downgrade on average. ### Module 3: Insights from Backbook Effect Analysis - Stylized path of rate shocks for funding cost projections: - In Y1 of stress: average funding costs of banks increase by 10 bps. - In Y2: funding costs increase by 30 bps. - In Y3: funding costs increase by 60 bps. - Rate shocks are applied to liabilities that reprice within one to three years using STE cash flow data, yielding funding cost projections for Y1, Y2, and Y3, and for repricing of new short-term liabilities in Y2 and Y3. - Pass-through and repricing assumptions: - Banks are allowed to pass-through rising funding costs to new loans in selected product markets (retail loans and loans to non-financial corporations where banks likely have pricing power). - Banks can only reprice new loans; variable rate contracts have fixed spreads above benchmark rates. - The STE data accounts for repricing of new short-term loans across Y1–Y3. - Minimum Hurdle Rate (MHR) and pass-through modeling: - Banks generate new loans at rates above the MHR; loans below MHR are not issued. - MHR includes funding and capital costs, risk costs, and operational costs attributed to the loan. - A pass-through rate of 100 percent of the weighted average funding costs on the repriceable loans is assumed for the stylized example. - Stylized MHR increases equal the funding cost increases: 10 bps in Y1, 30 bps in Y2, and 60 bps in Y3. - Frontbook vs backbook effect: - Frontbook effect = increase in interest income due to repricing of new loans. - Backbook effect = cumulative increase in funding costs (billion EUR) minus the increasing interest income from higher loan spreads over Y1 to Y3. - Exercise assesses what pass-through rate would be needed for the frontbook effect to offset the backbook effect. - Recapitalization and WACC impact (stylized): - The MHR approach can be used to calculate the impact of recapitalization on loan spreads via the impact on the weighted average cost of capital (WACC). - Formula provided: - ∆WACC = ∆r_L = ∆CET1 × [(RoE_t - (1-tax) × r_D) / ((1-tax))] - Assumption: tax = 25 percent for corporate profits in EA. ### Key quantitative findings on the backbook effect and funding feedback - Risk density (RD) defined as RWA/EAD (in percent) and asset-liability mismatch (ALM) defined as weighted average maturity of liabilities divided by weighted average maturity of repriceable loans (in percent) explain a large part of variation in the backbook effect. - Econometric results (Table 15) — contribution to ∆CET1: - RD coefficient = -1.49551*** (t-stat -4.32515) - ALM coefficient = 0.012794*** (t-stat 6.422979) - Constant C = 0.83231*** (t-stat 6.455153) - Adjusted R-squared = 0.608477 - S.E. of regression = 0.146822 - Prob(F-statistic) = 0.000003 - Aggregate and heterogeneity results: - Disregarding the backbook effect in solvency stress testing systematically and significantly underestimates the shock impact. - The average backbook effect amounts to about two-third of the entire funding cost effect. - To reduce the backbook effect to 0, the pass-through rate would have to be unrealistically high (above 500 percent). - The impact of a moderate average increase in funding costs on the CET1 ratio is economically significant with an average of 88 bps after three years. - The feedback-effect of funding costs on CET1 ratio is heterogeneous across banks and is higher for banks with: - (i) higher asset liability mismatch, - (ii) higher share of unsecured wholesale funding, - (iii) lower risk density, - (iv) lower volume of repriceable loans, - (v) less pricing power in loan markets (lower pass-through rate). ### Overall assessment - Resilience of large euro area banks has improved, but important vulnerabilities remain. - Capital buffers are in aggregate sizeable relative to immediate threats, but some banks are especially vulnerable to: - credit risk, and - market risks, including from risk premia decompression, basis risk, and valuation shocks. - IMF sensitivity-based reverse stress test results indicate that a moderate valuation shock on hard-to-value assets might deplete capital buffers of some large complex banks. - Low profitability is found in many banks across all business models, despite improving conjunctural conditions. - Diversity of the euro area banking sector: - Large but less complex internationally-active banks are more robust to stress scenarios and single-factor shocks (potential risk-sharing benefits from cross-border banking). - G-SIBs (with trading and capital markets-related activities) are disproportionately affected by market dislocation under severe market conditions. - Domestically-oriented banks are relatively more vulnerable to macroeconomic deteriorations. - Stress-test scope consideration: - The exercise is conducted at banking group consolidated level assuming free flow of capital and liquidity within each group; it does not capture intra-group frictions (regulated/unregulated, cross-border, cross-sectoral, upstream, downstream). - Prospective inclusion of significant investment firms under SSM supervision would necessitate and facilitate expanding the macroprudential stress testing framework to include them. ### Possible supervisory implications (Box 5) - Supervisory attention could prioritize credit risk in high-risk portfolios; prudential provisioning practices in some banks might need to be stepped up to build additional buffers. - Valuation effects from a sudden correction in asset prices, basis widening, and shocks to hard-to-value assets should be high priority. - Attention should be given to business risk from lower-than-expected business volumes, migration of client positions, compressed margins, and dynamic hedging. - Supervisors could focus attention on banks that feature a negative post-stress CBC; banks may need to increase their CBC, lengthen and stagger tenors of deposits from financial institutions, or improve the price risk-sensitivity of committed lines to customers. - Some banks might need to proactively strengthen their balance sheets (by increasing their capital base) to ensure reasonable funding costs as reliance on private funding increases; banks combining relatively low capitalization with low liquidity are especially vulnerable to combined shocks. *International Monetary Fund — EURO AREA POLICIES (excerpts from solvency stress test analysis).* ### Annex Table 4. Euro Area: Determinants of Deposit Rates in Low-Spread Country C1 ### Annex Table 4. Euro Area: Determinants of Deposit Rates in Low-Spread Country C1 ### Regressions: overview, sample, and estimation - Dependent variables: - Columns (1) and (2): deposit rate on new business from corporations. - Columns (3) and (4): deposit rate on new business from households. - Regressors include: l.new business (lagged deposit rate), g (GDP growth), inf (inflation rate), credit (private sector credit growth), ecb_pol (ECB main refinancing rate), libor_euro (3-month euro libor rate), swap (10-year swap rate), spread_swap (10-year sovereign bond yield over swap rate), and libor_usd (3-month euro libor rate). Lagged variables are denoted by L (t-1) and L2 (t-2). - Sample covers 2005Q1 through 2016Q4. - Estimation: Newey-West HAC-robust standard errors. - Significance notation: *** p<0.01, ** p<0.05, * p<0.1. ### Key coefficient estimates (column layout: (1) (2) for corporations; (3) (4) for households) - L.new business - (1): -0.321 (0.600) - (2): -0.009 (0.453) - (3): -0.375 (0.295) - (4): 0.020 (0.179) - L.g - (1): 0.058*** (0.014) - (2): 0.020 (0.021) - (3): 0.011 (0.010) - (4): 0.039 (0.031) - L.inf - (1): -0.037 (0.048) - (2): 0.099** (0.046) - (3): -0.015 (0.035) - (4): 0.118** (0.043) - L2.inf - (1): (not reported) - (2): -0.091* (0.049) - (3): (not reported) - (4): -0.074** (0.033) - L.credit - (1): -0.096* (0.050) - (2): -0.055 (0.053) - (3): 0.001 (0.025) - (4): -0.044 (0.031) - L.ecb_pol - (1): 0.486 (0.290) - (2): 0.503*** (0.121) - (3): 0.883*** (0.172) - (4): 0.615*** (0.094) - L2.ecb_pol - (1): (not reported) - (2): -0.719*** (0.138) - (3): (not reported) - (4): -0.277** (0.125) - L.libor_euro - (1): 0.711* (0.414) - (2): 0.745* (0.379) - (3): 0.341** (0.155) - (4): 0.305** (0.118) - L2.libor_euro - (1): 0.149 (0.210) - (2): (not reported) - (3): 0.009 (0.137) - (4): (not reported) - L.swap - (1): -0.215 (0.172) - (2): 0.403 (0.262) - (3): 0.009 (0.046) - (4): 0.018 (0.134) - L2.swap - (1): -0.529 (0.377) - (2): 0.047 (0.165) - (3): (not reported) - (4): (not reported) - L.spread_swap - (1): -0.005* (0.003) - (2): 0.003 (0.002) - (3): -0.003** (0.001) - (4): -0.002 (0.002) - L2.spread_swap - (1): (not reported) - (2): -0.004 (0.004) - (3): 0.002 (0.002) - (4): (not reported) - L.libor_usd - (1): 0.175*** (0.040) - (2): -0.140 (0.190) - (3): -0.087** (0.040) - (4): -0.268** (0.098) - L2.libor_usd - (1): (not reported) - (2): 0.324 (0.231) - (3): (not reported) - (4): 0.299** (0.131) - Constant - (1): 0.133 (0.240) - (2): 0.295 (0.282) - (3): 0.617*** (0.208) - (4): 0.347** (0.138) ### Model fit and sample size - Observations: - Columns (1) and (3): 47 - Columns (2) and (4): 46 - Adjusted R-squared: - (1): 0.979 - (2): 0.990 - (3): 0.987 - (4): 0.989 ### Notes on interpretation - Coefficient signs, magnitudes, and statistical significance are reported exactly as in the table with Newey-West HAC-robust standard errors. - The regressions include up to two lags (L and L2) for some macro and market variables where reported. *Source: Annex Table 4. Euro Area: Determinants of Deposit Rates in Low-Spread Country C1 (sample 2005Q1–2016Q4).* ### Annex Table 10. Euro Area: Effect ### Annex Table 10. Euro Area: Effect of Bank Solvency on Bank Retail Deposit Rates ### Model and sample - Dependent variable: customer deposit rate (bank specific). - Regressors include: g (GDP growth), u (unemployment rate), inf (inflation rate), libor_euro (3-month euro libor rate), ecb_pol (ECB main refinancing rate), spread_swap (10-year sovereign bond yield over swap rate), libor_usd_tbill (3-month usd libor over 3-month US T-bill rate), g_ea (GDP growth in the euro area), g ǀ g<0 (GDP growth negative), and tier 1 (tier 1 ratio). Lagged variables are denoted by L (t-1). - The sample covers 2005Q1 through 2016Q4. - Estimation: panel data with fixed effects at the country level (columns 1-9), and at the euro area level (columns 10 and 11). - Significance: *** p<0.01, ** p<0.05, * p<0.1. ### Key statistically significant coefficient estimates (lagged regressors, selected) - L.g: - Column C2: 0.121*** (SE 0.038) - Column C7: -0.017*** (SE 0.006) - L.u: - Column C3: 0.413*** (SE 0.133) - Column C6: 0.906*** (SE 0.283) - Column C7: -0.253*** (SE 0.053) - Column C9: -0.408** (SE 0.159) - Spec 2 (column 11): 0.028** (SE 0.014) - L.inf: - Column C4: 0.228*** (SE 0.068) - Column C7: 0.075** (SE 0.032) - Column C8: -0.196** (SE 0.056) - Column C9: 0.361** (SE 0.140) - L.libor_euro: - Column C1: 0.696* (SE 0.349) - Column C2: 0.209** (SE 0.102) - Column C7: 1.030*** (SE 0.181) - Column C8: 0.434*** (SE 0.130) - Column C9: 1.909*** (SE 0.624) - L.ecb_pol: - Column C2: 0.361*** (SE 0.105) - Column C4: 0.714*** (SE 0.160) - Column C6: 0.767** (SE 0.305) - Column C8: 0.746*** (SE 0.108) - Spec 1 (column 10): 0.867*** (SE 0.055) - Spec 2 (column 11): 0.821*** (SE 0.060) - L.spread_swap: - Column C1: 0.005** (SE 0.003) - Column C4: -0.003*** (SE 0.001) - Column C7: 0.002*** (SE 0.000) - L.libor usd_tbill: - Column C1: -0.698*** (SE 0.249) - Column C4: 0.354* (SE 0.196) - Column C7: -0.614*** (SE 0.089) - Column C9: -1.516*** (SE 0.381) - Spec 1 (column 10): -0.149** (SE 0.071) - L.g_ea: - Column C2: -0.157*** (SE 0.025) - Column C7: 0.112*** (SE 0.032) - Column C8: 0.159** (SE 0.068) - Spec 1 (column 10): -0.041* (SE 0.023) - Spec 2 (column 11): -0.054** (SE 0.022) - L.g ǀ g<0: - Column C2: -0.132*** (SE 0.043) - Column C6: -0.296* (SE 0.171) - Column C7: 0.038* (SE 0.021) - Tier1 effects included in select specs: - L.tier1 (column C7): -0.190** (SE 0.083) - L.tier1_sq (column C7): 0.006* (SE 0.003) ### Model fit, sample size, and identifiers - Observations by column: - C1: 77 - C2: 46 - C3: 133 - C4: 196 - C5: 47 - C6: 221 - C7: 83 - C8: 119 - C9: 113 - Spec 1 (column 10): 1,035 - Spec 2 (column 11): 988 - Adjusted R2 by column: - C1: 0.848 - C2: 0.982 - C3: 0.676 - C4: 0.755 - C5: 0.561 - C6: 0.476 - C7: 0.951 - C8: 0.901 - C9: 0.534 - Spec 1: 0.571 - Spec 2: 0.607 - Number of cnid (panel identifiers) by column: - C1: 2 - C2: 2 - C3: 6 - C4: 5 - C5: 1 - C6: 5 - C7: 2 - C8: 3 - C9: 3 - Spec 1: 29 - Spec 2: 29 ### Notable patterns and interpretations from the table - Several macro-financial lags display statistically significant associations with bank retail deposit rates across specifications, notably lagged euro LIBOR (L.libor_euro), ECB policy rate (L.ecb_pol), and lagged unemployment (L.u). - Funding-related international rate differentials (L.libor usd_tbill) show negative and significant coefficients in multiple specifications. - Bank solvency measures (tier1) enter in selected specifications: L.tier1 is negative and significant in one specification, while a positive tier1 squared term (L.tier1_sq) is also significant in the same specification, indicating potential non-linearities. - Panel-level (country vs euro area) specifications (Spec 1 and Spec 2) retain several significant macro-financial controls and show substantial sample size (1,035 and 988 observations) with adjusted R2 of 0.571 and 0.607 respectively. *Source: Annex Table 10. Euro Area: Effect of Bank Solvency on Bank Retail Deposit Rates. The sample covers 2005Q1 through 2016Q4. Estimation using panel data with fixed effects at the country level (columns 1-9), and at the euro area level (columns 10 and 11).*** p<0.01, ** p<0.05, * p<0.1.* ### 4. Risks and buffers ### 4. Risks and buffers ### Positions and risk factors assessed - Traded risk losses recognized the first year of stress (instantaneous shock). Net trading income from equity positions, debt instruments, and trading derivatives. - Interest income from defaulted loans: - Accrued on a net basis (2018 EBA methodology) for instantaneous shock. - No interest income accrual from defaulted loans in other TD/FSAP specifications. - Traded risk losses recognized the year that the shock hits (over the 3-year horizon), except for sensitivity tests (instantaneous shocks excluding low-for-long). Net trading income from equity positions, debt instruments, and trading derivatives. ### A. Banking Sector: Solvency Test — design and assumptions - Interest income and expenses: - Interest income from non-defaulting loans is estimated according to satellite models. - Interest expenses at the country level (by product and maturity) and bank level are linked to the scenario. - Interest expenses increase due to rising funding costs linked to banks’ funding structure and market shocks, with model-based pass-through on corporate and household loans. - Non-interest items and operating costs: - Net fee and commission income, non-interest income (e.g. insurance income, dividend income, other income), and operational expenses evolve with the scenario. - No change in business models (no rebalancing of portfolio is allowed). - Taxation and regulatory impact: - Tax Rate: 30 percent rate. - TD by ECB: No conversion of additional Tier 1 capital is assumed during the stress horizon. If banks’ capital ratio falls below regulatory minimum during the stress test horizon, no prompt corrective action is assumed. - TD by FSAP Team: The effects of the phase-out of no-longer-eligible additional Tier 1 and Tier 2 capital are included. No conversion of additional Tier 1 capital is assumed during the stress horizon. If banks’ capital ratio falls below regulatory minimum during the stress test horizon, no prompt corrective action is assumed. - Behavioral adjustments: - Static Balance Sheet: In line with 2018 EBA methodology. Maturing assets are replaced by exposures of the same type and risk. - Dynamic Balance Sheet: Credit demand shocks are included while credit supply effects are disallowed. EaD from off-balance sheet exposures increases under stress, reflecting higher use of undrawn credit and liquidity facilities. EaD evolves with structural foreign exchange risk. Maturing assets are replaced by exposures of the same type and risk. - Dividend policy: - Payout ratio according to 2018 EBA methodology (TD by ECB). - TD by FSAP Team: Dividend payout ratio linked to banks’ profits, capital ratios, and history, subject to constraints: - Floor set at 30 percent (positive profits); - CCB schedule for Common Equity Tier I; - Statistical analysis on banks’ dividend policies during stress episodes. ### A. Banking Sector: Solvency Test — parameter calibration and regulatory standards - Parameter calibration: - TD by ECB: Initial Point-in-Time (PiT) PD and LGD parameters for non-defaulted exposures for expected losses from 2016 EU-wide EBA stress test. Through-the-cycle (TTC) PD and LGD parameters for non-defaulted exposures for unexpected losses (i.e. RWAs) from 2016 EU-wide EBA stress test. Shifts to RWAs for IRB exposures. Historical PDs informed by NCAs’ submissions of default rates, Moody’s EDF rates, and PD proxies by Kamakura. - TD by FSAP Team: Initial regulatory PD and LGD parameters (hybrid PiT and TTC models) using COREP supervisory data by geographic and portfolio breakdown on the obligor pool. Calculations performed to extract PD and LGD for non-defaulted exposures using information related to gross defaulted exposures (09.01 and 09.02 templates) and breakdown by obligor grade (08.02). Shifts to IRB and STA exposures. Historical PDs informed by Moody’s EDF proxies, Merton-model approach for sovereign spreads, and bank-specific PDs from Pillar 3 disclosures. - Regulatory standards: - Capital definition according to national implementation of Capital Requirements Directive (CRD) IV rulebook, including CET1, Tier 1, and total CAR. - Capital components that are no longer eligible for additional Tier 1 and Tier 2 capital components follow Basel III transitional path. - The CET1 hurdle rate consisting of a 4.5 percent Pillar 1 requirement, 0.625 percent capital conservation (CCB) buffer, and phased-in bank-specific G-SII. - For reverse sensitivity test the CET1 hurdle rate also includes Pillar 2 requirement and systemic risk buffer (SRB). ### Reporting format for solvency results - Output presentation (aggregate euro area banking system and by type of bank — G-SIBs, Less Complex Large Internationally Active Banks, Relatively Smaller Domestically-Oriented Banks): - Distribution of capital ratios under baseline/adverse scenario (box plots); - Contribution to profitability and capital depletion by driver; - Average CET1, CAR, and Tier 1 leverage ratio. ### B. Banking Sector: Liquidity Test — perimeter and data - Institutions included: - 29 banks on the consolidated basis. - Market share: Over 70 percent of total banking sector assets. - Supervisory data: ALMM Maturity Ladder Template STE. Consolidated basis. Banks grouped by business model. - Baseline date: September 30, 2017. ### B. Banking Sector: Liquidity Test — channels, scenarios, and feedbacks - Methodology: - Cash flow-based analysis using contractual and behavioral (where available) cash flow data for significant currencies with assumptions about combined interaction of funding and market liquidity and different degrees of central bank support. - LCR and NSFR analysis using granular data templates. - Liquidity stock (maturity transformation) analysis using NFSR data. - Five days collateral freeze scenario assuming that collateral received is not available for rehypothecation. - Feedback loops and links with solvency analysis: - Exploratory scenario: Solvency-Funding cost loop (standardized shock of 100 bps in funding costs). - Financial-Macro feedback loop: Changes in credit cost and volumes of new loans lead to changes in key macro variables (GDP, Unemployment etc.). - Sensitivity analysis perimeter and types: - LCR distribution and volatility across banks and significant currencies. - NSFR distribution across banks. ### B. Banking Sector: Liquidity Test — tail shocks and scenario design - Size and types of shocks: - Baseline: business as usual (as reported by banks under normal market conditions). Behavioral assumptions: all maturing liabilities are rolled-over. - 5-day collateral freeze scenario (due to cyber-risk related event at CCP). - 1-month intermediate/severe market stress scenario: higher run-off rates on unsecured wholesale funding (incl. FX swaps), and undrawn committed credit/liquidity lines on top of the mild stress scenario. - 1-month severe combined (market/idiosyncratic) scenario. - 3-month intermediate/severe market stress scenario: higher run-off rates on secured wholesale funding (particularly FX swaps) on top of the intermediate stress scenario. - 3-month severe combined (market/idiosyncratic) scenario. - Each scenario provides for three approaches to the CBC with decreasing reliance on the CB and increasing focus on market liquidity (e.g. asset liquidation, asset encumbrance and collateral swaps). - All scenarios are EUR based (acc. across all currencies) and USD based. - Total number of scenarios: 40 (four sets of embedded scenarios of increasing severity). - Liquidity concentration test: loss of funding from the largest providers. ### B. Banking Sector: Liquidity Test — standards, fail criteria, and reporting - Regulatory and market-based standards and parameters: - Threshold for cash flow-based analysis: net cumulative funding gap falls below zero. - Threshold for LCRs set to 100 percent. - Fail criteria: - Fail criteria for cash flow-based liquidity analysis in foreign currencies: decrease of CBC in USD below 0. - Fail criteria for cash flow-based analysis across all currencies: CBC across all currencies below 0. - Reporting format for results: - Number of banks with negative net cumulative funding gaps in EUR (acc. across all currencies) and USD; - Aggregate negative cumulated counterbalancing capacity. *International Monetary Fund — cr18228, 4. Risks and buffers* --- _Source: https://www.imf.org/-/media/files/publications/cr/2018/cr18228.pdf_